Trouble with multi-dimensional analysis and home installation

system · November 20, 2007, 1:38pm

We are two students at LTH in Lund, Sweden who are using Anybody
v3.0.0 to do a biomechanical analysis of a cyclist as part of our
master thesis project. Unfortunately we have run into a little
difficulty and hope you can help us with a few points. We have
adapted the Bike Model 3D program from the repository to run
parameter and optimisation studies so that we can investigate the
effects of varying saddle height, saddle position and pedal arm
width, in the hope of comparing these results to real-world results
from test subjects.

Our first and greatest problem is that our program is prone to
unexpected, but reproducible, crashes that simply halt the analysis
with no error messages given or explanations as to the cause. The
program gives no compilation errors when loaded, and the crashes seem
to be mainly caused by the number of steps we make the program
investigate. Our most recent crash, for example, involved an n =
{4,4,4} step analysis, given a certain set starting value and
variable values for each parameter. A previously run n = {2,2,2} step
analysis, with identical start values and variable values, ran
through to its conclusion with no crashes. We found we could reduce
the number of crashes by using the simple muscle model as opposed to
the 3E model, but they still occur, thus far, every time we attempt an
analysis of more than around 40 steps in total.

Our second problem is that we cannot export our results to an external
browser. Given our need for a 4-dimensional graph to view our results
satisfactorily, the ability to export the results is vitally
important for us. We realise that it is possible, but are having
difficulties implemeting the commands found under the Export folder
in the properties meny of the 2D/3D Chart. Any pointers you could
give us on how to do this properly would be greatly appreciated!

Thirdly, I have attempted to install Anybody on my home computer, with
which we could perform these analyses much faster than on campus, but
am having problems with the student license dongle installation. I am
running Vista 64-bit, which I realise isn’t ideal for the moment, but
I have managed to install all the relevant parts of Anybody itself
and the License Manager using Windows XP SP2 compatibility mode. When
I try to open Anybody, the License window appears, and when I
press “Dongle Bound License”, a screen pops up telling me to insert
the dongle. Unfortunately, this happens despite the fact that the
dongle is already inserted in the USB drive with a constant red
light, and the HASP License Manager is active (even when in XP SP2
compatability mode).

We would greatly appreciate any help you can give us! The school PC
we are trying to run on is a P4 1.8GHz with 512mb RAM with WinXP; is
this a factor in the crashes? If so, running it on my home computer
(Core 2 Duo 2.66GHz, 4gb RAM) is an even more pressing necessity,
given that Vista isn’t a limiting factor on its own.

Hope you can help us! Thanks in advance. We can of course send you our
program if you need it.

Henrik SjÃ¶stedt
Tobias Lestrell

system · November 20, 2007, 9:34pm

Hi Henrik and Tobias

Welcome to the group, it sounds as an interesting problem.

From your description it sounds as you may have a recruitment problem, but
it is difficult to say from the description.

To find out if it is recruitment related please try to rerun the parameter
study but only run the kinematical analysis.

If the error is related to recruitment and if the error is reproducible my
suggestion would be to try to locate which parameter set, in the parameters
study are causing this error. I think that it may be the combination of the
parameters rather than the number of steps in the parameter study which are
causing the crash. So if you try running the parameter study again and then
notice for which parameter set it crash? Then subsequently try to rerun the
model with this set of parameters only and see if the error can be
reproduced. If this is that case, then investigate the parameters and if
possible evaluate if there could be any apparent reason for the problem. If
this seems not to be the case you could try to change the recruitment solver
it may possible solve the problem.

Please read more about the different solvers in the tutorial

http://www.anybodytech.com/691.0.html

There are two Linear algorithms to choose between MinMaxSimplex and
MinMaxNRSimplex , if you have tried only one of these, please try the other
one and see if this solves the problem, the results are identical.

In some cases if the problem is ill conditioned, it helps to add a small
amount of gravity in the zero direction of the gravity vector, so something
like Gravity={0.001,-9.81,0.001}; may also help solve recruitment problems
if this is the root of the problem, it will not change the solution.

If the problems persist, please try to find the parameter set causing the
problem and then upload this model to the file section of the group, then we
will try to debug it, please also write which version of AnyBody you are
using.

About the output, please try to insert an AnyOutPutFile object in your study
this will let you export as many variables you like to a text file.

I am uncertain about the installation problem; i will pass this question on
the development team.

I hope this helped you solved the problem; otherwise please do not hesitate
to write again.

Best regards

SÃ¸ren, AnyBody Support

From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of vtveteran
Sent: 20 November 2007 13:39
To: anyscript@yahoogroups.com
Subject: [AnyScript] Trouble with multi-dimensional analysis and home
installation

We are two students at LTH in Lund, Sweden who are using Anybody
v3.0.0 to do a biomechanical analysis of a cyclist as part of our
master thesis project. Unfortunately we have run into a little
difficulty and hope you can help us with a few points. We have
adapted the Bike Model 3D program from the repository to run
parameter and optimisation studies so that we can investigate the
effects of varying saddle height, saddle position and pedal arm
width, in the hope of comparing these results to real-world results
from test subjects.

Our first and greatest problem is that our program is prone to
unexpected, but reproducible, crashes that simply halt the analysis
with no error messages given or explanations as to the cause. The
program gives no compilation errors when loaded, and the crashes seem
to be mainly caused by the number of steps we make the program
investigate. Our most recent crash, for example, involved an n =
{4,4,4} step analysis, given a certain set starting value and
variable values for each parameter. A previously run n = {2,2,2} step
analysis, with identical start values and variable values, ran
through to its conclusion with no crashes. We found we could reduce
the number of crashes by using the simple muscle model as opposed to
the 3E model, but they still occur, thus far, every time we attempt an
analysis of more than around 40 steps in total.

Our second problem is that we cannot export our results to an external
browser. Given our need for a 4-dimensional graph to view our results
satisfactorily, the ability to export the results is vitally
important for us. We realise that it is possible, but are having
difficulties implemeting the commands found under the Export folder
in the properties meny of the 2D/3D Chart. Any pointers you could
give us on how to do this properly would be greatly appreciated!

Thirdly, I have attempted to install Anybody on my home computer, with
which we could perform these analyses much faster than on campus, but
am having problems with the student license dongle installation. I am
running Vista 64-bit, which I realise isn’t ideal for the moment, but
I have managed to install all the relevant parts of Anybody itself
and the License Manager using Windows XP SP2 compatibility mode. When
I try to open Anybody, the License window appears, and when I
press “Dongle Bound License”, a screen pops up telling me to insert
the dongle. Unfortunately, this happens despite the fact that the
dongle is already inserted in the USB drive with a constant red
light, and the HASP License Manager is active (even when in XP SP2
compatability mode).

We would greatly appreciate any help you can give us! The school PC
we are trying to run on is a P4 1.8GHz with 512mb RAM with WinXP; is
this a factor in the crashes? If so, running it on my home computer
(Core 2 Duo 2.66GHz, 4gb RAM) is an even more pressing necessity,
given that Vista isn’t a limiting factor on its own.

Hope you can help us! Thanks in advance. We can of course send you our
program if you need it.

Henrik SjÃ¶stedt
Tobias Lestrell

[Non-text portions of this message have been removed]

system · November 21, 2007, 10:39am

Hi Henrik and Tobias,

I am writing you regarding the Vista 64 issue. We have only little
experience with Vista and in particular Vista 64 yet. We are currently
preparing to support Vista 32 better in future versions of AnyBody.

I would like to help you as much as I can, but you will have to help me too
then, i.e., to gain experience into the problem. I would very much
appreciate this assistance from you.

If you are ready for this I suggest that we take the conversation off the
group while experimenting.

Best regards,

Michael

Michael Damsgaard, AnyBody Support

<http://www.anybodytech.com> www.anybodytech.com

From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of AnyBody Support
Sent: Tuesday, November 20, 2007 21:35
To: anyscript@yahoogroups.com
Subject: RE: [AnyScript] Trouble with multi-dimensional analysis and home
installation

Hi Henrik and Tobias

Welcome to the group, it sounds as an interesting problem.

From your description it sounds as you may have a recruitment problem, but
it is difficult to say from the description.

To find out if it is recruitment related please try to rerun the parameter
study but only run the kinematical analysis.

If the error is related to recruitment and if the error is reproducible my
suggestion would be to try to locate which parameter set, in the parameters
study are causing this error. I think that it may be the combination of the
parameters rather than the number of steps in the parameter study which are
causing the crash. So if you try running the parameter study again and then
notice for which parameter set it crash? Then subsequently try to rerun the
model with this set of parameters only and see if the error can be
reproduced. If this is that case, then investigate the parameters and if
possible evaluate if there could be any apparent reason for the problem. If
this seems not to be the case you could try to change the recruitment solver
it may possible solve the problem.

Please read more about the different solvers in the tutorial

http://www.anybodyt <http://www.anybodytech.com/691.0.html>
ech.com/691.0.html

There are two Linear algorithms to choose between MinMaxSimplex and
MinMaxNRSimplex , if you have tried only one of these, please try the other
one and see if this solves the problem, the results are identical.

In some cases if the problem is ill conditioned, it helps to add a small
amount of gravity in the zero direction of the gravity vector, so something
like Gravity={0.001,-9.81,0.001}; may also help solve recruitment problems
if this is the root of the problem, it will not change the solution.

If the problems persist, please try to find the parameter set causing the
problem and then upload this model to the file section of the group, then we
will try to debug it, please also write which version of AnyBody you are
using.

About the output, please try to insert an AnyOutPutFile object in your study
this will let you export as many variables you like to a text file.

I am uncertain about the installation problem; i will pass this question on
the development team.

I hope this helped you solved the problem; otherwise please do not hesitate
to write again.

Best regards

SÃ¸ren, AnyBody Support

From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
[mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com] On
Behalf
Of vtveteran
Sent: 20 November 2007 13:39
To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
Subject: [AnyScript] Trouble with multi-dimensional analysis and home
installation

We are two students at LTH in Lund, Sweden who are using Anybody
v3.0.0 to do a biomechanical analysis of a cyclist as part of our
master thesis project. Unfortunately we have run into a little
difficulty and hope you can help us with a few points. We have
adapted the Bike Model 3D program from the repository to run
parameter and optimisation studies so that we can investigate the
effects of varying saddle height, saddle position and pedal arm
width, in the hope of comparing these results to real-world results
from test subjects.

Our first and greatest problem is that our program is prone to
unexpected, but reproducible, crashes that simply halt the analysis
with no error messages given or explanations as to the cause. The
program gives no compilation errors when loaded, and the crashes seem
to be mainly caused by the number of steps we make the program
investigate. Our most recent crash, for example, involved an n =
{4,4,4} step analysis, given a certain set starting value and
variable values for each parameter. A previously run n = {2,2,2} step
analysis, with identical start values and variable values, ran
through to its conclusion with no crashes. We found we could reduce
the number of crashes by using the simple muscle model as opposed to
the 3E model, but they still occur, thus far, every time we attempt an
analysis of more than around 40 steps in total.

Our second problem is that we cannot export our results to an external
browser. Given our need for a 4-dimensional graph to view our results
satisfactorily, the ability to export the results is vitally
important for us. We realise that it is possible, but are having
difficulties implemeting the commands found under the Export folder
in the properties meny of the 2D/3D Chart. Any pointers you could
give us on how to do this properly would be greatly appreciated!

Thirdly, I have attempted to install Anybody on my home computer, with
which we could perform these analyses much faster than on campus, but
am having problems with the student license dongle installation. I am
running Vista 64-bit, which I realise isn’t ideal for the moment, but
I have managed to install all the relevant parts of Anybody itself
and the License Manager using Windows XP SP2 compatibility mode. When
I try to open Anybody, the License window appears, and when I
press “Dongle Bound License”, a screen pops up telling me to insert
the dongle. Unfortunately, this happens despite the fact that the
dongle is already inserted in the USB drive with a constant red
light, and the HASP License Manager is active (even when in XP SP2
compatability mode).

We would greatly appreciate any help you can give us! The school PC
we are trying to run on is a P4 1.8GHz with 512mb RAM with WinXP; is
this a factor in the crashes? If so, running it on my home computer
(Core 2 Duo 2.66GHz, 4gb RAM) is an even more pressing necessity,
given that Vista isn’t a limiting factor on its own.

Hope you can help us! Thanks in advance. We can of course send you our
program if you need it.

Henrik SjÃ¶stedt
Tobias Lestrell

[Non-text portions of this message have been removed]

system · November 22, 2007, 11:00am

Hi SÃ¸ren and Michael,

Thanks a lot for your advice! Changing the linear algorithm to
MinMaxSimplex appears to have done the trick, and we’ve run analyses
up to 125 steps without issue. We did leave a {10,10,10} analysis
running overnight that crashed, but this feels like a different issue
because instead of the analysis simply halting mid-stride, Windows
flagged the program as crashed and asked if we wanted to send a
report, which hasn’t been the case before. We also managed to sort
out a working output file, though we do wonder if there is any way
Excel can recognise the different columns on its own.

As concerns the Vista64 issue, I’ll be glad to help out, though I’m
obviously at uni most of the day, away from my home PC. The issue I’m
having at the moment I presume is with the HASP License Manager, as
opposed to Anybody itself.

Thanks for all the help!

Henrik SjÃ¶stedt
Tobias Lestrell

— In anyscript@yahoogroups.com, “AnyBody Support” <support@…>
wrote:
>
> Hi Henrik and Tobias,
>
>
>
> I am writing you regarding the Vista 64 issue. We have only little
> experience with Vista and in particular Vista 64 yet. We are
currently
> preparing to support Vista 32 better in future versions of AnyBody.
>
> I would like to help you as much as I can, but you will have to
help me too
> then, i.e., to gain experience into the problem. I would very much
> appreciate this assistance from you.
>
> If you are ready for this I suggest that we take the conversation
off the
> group while experimenting.
>
>
>
> Best regards,
>
>
>
> Michael
>
> ------------------------------
>
> Michael Damsgaard, AnyBody Support
>
> <http://www.anybodytech.com> www.anybodytech.com
>
>
>
>
>
> _____
>
> From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com]
On Behalf
> Of AnyBody Support
> Sent: Tuesday, November 20, 2007 21:35
> To: anyscript@yahoogroups.com
> Subject: RE: [AnyScript] Trouble with multi-dimensional analysis
and home
> installation
>
>
>
> Hi Henrik and Tobias
>
> Welcome to the group, it sounds as an interesting problem.
>
> From your description it sounds as you may have a recruitment
problem, but
> it is difficult to say from the description.
>
> To find out if it is recruitment related please try to rerun the
parameter
> study but only run the kinematical analysis.
>
> If the error is related to recruitment and if the error is
reproducible my
> suggestion would be to try to locate which parameter set, in the
parameters
> study are causing this error. I think that it may be the
combination of the
> parameters rather than the number of steps in the parameter study
which are
> causing the crash. So if you try running the parameter study again
and then
> notice for which parameter set it crash? Then subsequently try to
rerun the
> model with this set of parameters only and see if the error can be
> reproduced. If this is that case, then investigate the parameters
and if
> possible evaluate if there could be any apparent reason for the
problem. If
> this seems not to be the case you could try to change the
recruitment solver
> it may possible solve the problem.
>
> Please read more about the different solvers in the tutorial
>
> http://www.anybodyt <http://www.anybodytech.com/691.0.html>
> ech.com/691.0.html
>
> There are two Linear algorithms to choose between MinMaxSimplex and
> MinMaxNRSimplex , if you have tried only one of these, please try
the other
> one and see if this solves the problem, the results are identical.
>
> In some cases if the problem is ill conditioned, it helps to add a
small
> amount of gravity in the zero direction of the gravity vector, so
something
> like Gravity={0.001,-9.81,0.001}; may also help solve recruitment
problems
> if this is the root of the problem, it will not change the solution.
>
> If the problems persist, please try to find the parameter set
causing the
> problem and then upload this model to the file section of the
group, then we
> will try to debug it, please also write which version of AnyBody
you are
> using.
>
> About the output, please try to insert an AnyOutPutFile object in
your study
> this will let you export as many variables you like to a text file.
>
> I am uncertain about the installation problem; i will pass this
question on
> the development team.
>
> I hope this helped you solved the problem; otherwise please do not
hesitate
> to write again.
>
> Best regards
>
> SÃ¸ren, AnyBody Support
>
> _____
>
> From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com] On
> Behalf
> Of vtveteran
> Sent: 20 November 2007 13:39
> To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> Subject: [AnyScript] Trouble with multi-dimensional analysis and
home
> installation
>
> We are two students at LTH in Lund, Sweden who are using Anybody
> v3.0.0 to do a biomechanical analysis of a cyclist as part of our
> master thesis project. Unfortunately we have run into a little
> difficulty and hope you can help us with a few points. We have
> adapted the Bike Model 3D program from the repository to run
> parameter and optimisation studies so that we can investigate the
> effects of varying saddle height, saddle position and pedal arm
> width, in the hope of comparing these results to real-world results
> from test subjects.
>
> Our first and greatest problem is that our program is prone to
> unexpected, but reproducible, crashes that simply halt the analysis
> with no error messages given or explanations as to the cause. The
> program gives no compilation errors when loaded, and the crashes
seem
> to be mainly caused by the number of steps we make the program
> investigate. Our most recent crash, for example, involved an n =
> {4,4,4} step analysis, given a certain set starting value and
> variable values for each parameter. A previously run n = {2,2,2}
step
> analysis, with identical start values and variable values, ran
> through to its conclusion with no crashes. We found we could reduce
> the number of crashes by using the simple muscle model as opposed
to
> the 3E model, but they still occur, thus far, every time we attempt
an
> analysis of more than around 40 steps in total.
>
> Our second problem is that we cannot export our results to an
external
> browser. Given our need for a 4-dimensional graph to view our
results
> satisfactorily, the ability to export the results is vitally
> important for us. We realise that it is possible, but are having
> difficulties implemeting the commands found under the Export folder
> in the properties meny of the 2D/3D Chart. Any pointers you could
> give us on how to do this properly would be greatly appreciated!
>
> Thirdly, I have attempted to install Anybody on my home computer,
with
> which we could perform these analyses much faster than on campus,
but
> am having problems with the student license dongle installation. I
am
> running Vista 64-bit, which I realise isn’t ideal for the moment,
but
> I have managed to install all the relevant parts of Anybody itself
> and the License Manager using Windows XP SP2 compatibility mode.
When
> I try to open Anybody, the License window appears, and when I
> press “Dongle Bound License”, a screen pops up telling me to insert
> the dongle. Unfortunately, this happens despite the fact that the
> dongle is already inserted in the USB drive with a constant red
> light, and the HASP License Manager is active (even when in XP SP2
> compatability mode).
>
> We would greatly appreciate any help you can give us! The school PC
> we are trying to run on is a P4 1.8GHz with 512mb RAM with WinXP;
is
> this a factor in the crashes? If so, running it on my home computer
> (Core 2 Duo 2.66GHz, 4gb RAM) is an even more pressing necessity,
> given that Vista isn’t a limiting factor on its own.
>
> Hope you can help us! Thanks in advance. We can of course send you
our
> program if you need it.
>
> Henrik SjÃ¶stedt
> Tobias Lestrell
>
> [Non-text portions of this message have been removed]
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

system · November 22, 2007, 11:28am

Hi again guys,

a slightly simpler problem this time. Our examinator Ingrid Svensson
at LTH, Lund, can’t find any of the student license keys in the
package you sent to her, so we can’t install the student licenses on
any of the school computers either. As it is, we’re excited to find
out what happens when our Trial versions expire on the 31st November
(that’s the date we were given when we signed up on 31st October :-
))!!

Just to clarify, this isn’t really relevant to the issues I’m having
with home installation, because at home the dongle isn’t even being
recognised.

Thanks again for all your help!

— In anyscript@yahoogroups.com, “vtveteran” <henriksjostedt@…>
wrote:
>
> Hi SÃ¸ren and Michael,
>
> Thanks a lot for your advice! Changing the linear algorithm to
> MinMaxSimplex appears to have done the trick, and we’ve run
analyses
> up to 125 steps without issue. We did leave a {10,10,10} analysis
> running overnight that crashed, but this feels like a different
issue
> because instead of the analysis simply halting mid-stride, Windows
> flagged the program as crashed and asked if we wanted to send a
> report, which hasn’t been the case before. We also managed to sort
> out a working output file, though we do wonder if there is any way
> Excel can recognise the different columns on its own.
>
> As concerns the Vista64 issue, I’ll be glad to help out, though I’m
> obviously at uni most of the day, away from my home PC. The issue
I’m
> having at the moment I presume is with the HASP License Manager, as
> opposed to Anybody itself.
>
> Thanks for all the help!
>
> Henrik SjÃ¶stedt
> Tobias Lestrell
>
>
>
> — In anyscript@yahoogroups.com, “AnyBody Support” <support@>
> wrote:
> >
> > Hi Henrik and Tobias,
> >
> >
> >
> > I am writing you regarding the Vista 64 issue. We have only little
> > experience with Vista and in particular Vista 64 yet. We are
> currently
> > preparing to support Vista 32 better in future versions of
AnyBody.
> >
> > I would like to help you as much as I can, but you will have to
> help me too
> > then, i.e., to gain experience into the problem. I would very much
> > appreciate this assistance from you.
> >
> > If you are ready for this I suggest that we take the conversation
> off the
> > group while experimenting.
> >
> >
> >
> > Best regards,
> >
> >
> >
> > Michael
> >
> > ------------------------------
> >
> > Michael Damsgaard, AnyBody Support
> >
> > <http://www.anybodytech.com> www.anybodytech.com
> >
> >
> >
> >
> >
> > _____
> >
> > From: anyscript@yahoogroups.com
[mailto:anyscript@yahoogroups.com]
> On Behalf
> > Of AnyBody Support
> > Sent: Tuesday, November 20, 2007 21:35
> > To: anyscript@yahoogroups.com
> > Subject: RE: [AnyScript] Trouble with multi-dimensional analysis
> and home
> > installation
> >
> >
> >
> > Hi Henrik and Tobias
> >
> > Welcome to the group, it sounds as an interesting problem.
> >
> > From your description it sounds as you may have a recruitment
> problem, but
> > it is difficult to say from the description.
> >
> > To find out if it is recruitment related please try to rerun the
> parameter
> > study but only run the kinematical analysis.
> >
> > If the error is related to recruitment and if the error is
> reproducible my
> > suggestion would be to try to locate which parameter set, in the
> parameters
> > study are causing this error. I think that it may be the
> combination of the
> > parameters rather than the number of steps in the parameter study
> which are
> > causing the crash. So if you try running the parameter study
again
> and then
> > notice for which parameter set it crash? Then subsequently try to
> rerun the
> > model with this set of parameters only and see if the error can be
> > reproduced. If this is that case, then investigate the parameters
> and if
> > possible evaluate if there could be any apparent reason for the
> problem. If
> > this seems not to be the case you could try to change the
> recruitment solver
> > it may possible solve the problem.
> >
> > Please read more about the different solvers in the tutorial
> >
> > http://www.anybodyt <http://www.anybodytech.com/691.0.html>
> > ech.com/691.0.html
> >
> > There are two Linear algorithms to choose between MinMaxSimplex
and
> > MinMaxNRSimplex , if you have tried only one of these, please try
> the other
> > one and see if this solves the problem, the results are
identical.
> >
> > In some cases if the problem is ill conditioned, it helps to add
a
> small
> > amount of gravity in the zero direction of the gravity vector, so
> something
> > like Gravity={0.001,-9.81,0.001}; may also help solve recruitment
> problems
> > if this is the root of the problem, it will not change the
solution.
> >
> > If the problems persist, please try to find the parameter set
> causing the
> > problem and then upload this model to the file section of the
> group, then we
> > will try to debug it, please also write which version of AnyBody
> you are
> > using.
> >
> > About the output, please try to insert an AnyOutPutFile object in
> your study
> > this will let you export as many variables you like to a text
file.
> >
> > I am uncertain about the installation problem; i will pass this
> question on
> > the development team.
> >
> > I hope this helped you solved the problem; otherwise please do
not
> hesitate
> > to write again.
> >
> > Best regards
> >
> > SÃ¸ren, AnyBody Support
> >
> > _____
> >
> > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com
> > [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com] On
> > Behalf
> > Of vtveteran
> > Sent: 20 November 2007 13:39
> > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> > Subject: [AnyScript] Trouble with multi-dimensional analysis and
> home
> > installation
> >
> > We are two students at LTH in Lund, Sweden who are using Anybody
> > v3.0.0 to do a biomechanical analysis of a cyclist as part of our
> > master thesis project. Unfortunately we have run into a little
> > difficulty and hope you can help us with a few points. We have
> > adapted the Bike Model 3D program from the repository to run
> > parameter and optimisation studies so that we can investigate the
> > effects of varying saddle height, saddle position and pedal arm
> > width, in the hope of comparing these results to real-world
results
> > from test subjects.
> >
> > Our first and greatest problem is that our program is prone to
> > unexpected, but reproducible, crashes that simply halt the
analysis
> > with no error messages given or explanations as to the cause. The
> > program gives no compilation errors when loaded, and the crashes
> seem
> > to be mainly caused by the number of steps we make the program
> > investigate. Our most recent crash, for example, involved an n =
> > {4,4,4} step analysis, given a certain set starting value and
> > variable values for each parameter. A previously run n = {2,2,2}
> step
> > analysis, with identical start values and variable values, ran
> > through to its conclusion with no crashes. We found we could
reduce
> > the number of crashes by using the simple muscle model as opposed
> to
> > the 3E model, but they still occur, thus far, every time we
attempt
> an
> > analysis of more than around 40 steps in total.
> >
> > Our second problem is that we cannot export our results to an
> external
> > browser. Given our need for a 4-dimensional graph to view our
> results
> > satisfactorily, the ability to export the results is vitally
> > important for us. We realise that it is possible, but are having
> > difficulties implemeting the commands found under the Export
folder
> > in the properties meny of the 2D/3D Chart. Any pointers you could
> > give us on how to do this properly would be greatly appreciated!
> >
> > Thirdly, I have attempted to install Anybody on my home computer,
> with
> > which we could perform these analyses much faster than on campus,
> but
> > am having problems with the student license dongle installation.
I
> am
> > running Vista 64-bit, which I realise isn’t ideal for the moment,
> but
> > I have managed to install all the relevant parts of Anybody
itself
> > and the License Manager using Windows XP SP2 compatibility mode.
> When
> > I try to open Anybody, the License window appears, and when I
> > press “Dongle Bound License”, a screen pops up telling me to
insert
> > the dongle. Unfortunately, this happens despite the fact that the
> > dongle is already inserted in the USB drive with a constant red
> > light, and the HASP License Manager is active (even when in XP
SP2
> > compatability mode).
> >
> > We would greatly appreciate any help you can give us! The school
PC
> > we are trying to run on is a P4 1.8GHz with 512mb RAM with WinXP;
> is
> > this a factor in the crashes? If so, running it on my home
computer
> > (Core 2 Duo 2.66GHz, 4gb RAM) is an even more pressing necessity,
> > given that Vista isn’t a limiting factor on its own.
> >
> > Hope you can help us! Thanks in advance. We can of course send
you
> our
> > program if you need it.
> >
> > Henrik SjÃ¶stedt
> > Tobias Lestrell
> >
> > [Non-text portions of this message have been removed]
> >
> >
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
>

system · November 22, 2007, 1:04pm

Hi Henrik and Tobias

I am glad that it works now.

In the AnyOutPutFile object there are various settings available please see
more about this in the reference manual. By default each column is separated
by a “,”. So if you in the “Data” menu of Excel choose “import external
data” this will let you choose to have a comma as separation sign. If you
prefer to have another sign as for the separation this is also possible.

Best regards

SÃ¸ren, AnyBody Support

From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of vtveteran
Sent: 22 November 2007 11:01
To: anyscript@yahoogroups.com
Subject: [AnyScript] Re: Trouble with multi-dimensional analysis and home
installation

Hi SÃ¸ren and Michael,

Thanks a lot for your advice! Changing the linear algorithm to
MinMaxSimplex appears to have done the trick, and we’ve run analyses
up to 125 steps without issue. We did leave a {10,10,10} analysis
running overnight that crashed, but this feels like a different issue
because instead of the analysis simply halting mid-stride, Windows
flagged the program as crashed and asked if we wanted to send a
report, which hasn’t been the case before. We also managed to sort
out a working output file, though we do wonder if there is any way
Excel can recognise the different columns on its own.

As concerns the Vista64 issue, I’ll be glad to help out, though I’m
obviously at uni most of the day, away from my home PC. The issue I’m
having at the moment I presume is with the HASP License Manager, as
opposed to Anybody itself.

Thanks for all the help!

Henrik SjÃ¶stedt
Tobias Lestrell

— In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
“AnyBody Support” <support@…>
wrote:
>
> Hi Henrik and Tobias,
>
>
>
> I am writing you regarding the Vista 64 issue. We have only little
> experience with Vista and in particular Vista 64 yet. We are
currently
> preparing to support Vista 32 better in future versions of AnyBody.
>
> I would like to help you as much as I can, but you will have to
help me too
> then, i.e., to gain experience into the problem. I would very much
> appreciate this assistance from you.
>
> If you are ready for this I suggest that we take the conversation
off the
> group while experimenting.
>
>
>
> Best regards,
>
>
>
> Michael
>
> ------------------------------
>
> Michael Damsgaard, AnyBody Support
>
> <http://www.anybodyt <http://www.anybodytech.com> ech.com>
www.anybodytech.com
>
>
>
>
>
> _____
>
> From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
[mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com]
On Behalf
> Of AnyBody Support
> Sent: Tuesday, November 20, 2007 21:35
> To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> Subject: RE: [AnyScript] Trouble with multi-dimensional analysis
and home
> installation
>
>
>
> Hi Henrik and Tobias
>
> Welcome to the group, it sounds as an interesting problem.
>
> From your description it sounds as you may have a recruitment
problem, but
> it is difficult to say from the description.
>
> To find out if it is recruitment related please try to rerun the
parameter
> study but only run the kinematical analysis.
>
> If the error is related to recruitment and if the error is
reproducible my
> suggestion would be to try to locate which parameter set, in the
parameters
> study are causing this error. I think that it may be the
combination of the
> parameters rather than the number of steps in the parameter study
which are
> causing the crash. So if you try running the parameter study again
and then
> notice for which parameter set it crash? Then subsequently try to
rerun the
> model with this set of parameters only and see if the error can be
> reproduced. If this is that case, then investigate the parameters
and if
> possible evaluate if there could be any apparent reason for the
problem. If
> this seems not to be the case you could try to change the
recruitment solver
> it may possible solve the problem.
>
> Please read more about the different solvers in the tutorial
>
> http://www.anybodyt <http://www.anybodyt
<http://www.anybodytech.com/691.0.html> ech.com/691.0.html>
> ech.com/691.0.html
>
> There are two Linear algorithms to choose between MinMaxSimplex and
> MinMaxNRSimplex , if you have tried only one of these, please try
the other
> one and see if this solves the problem, the results are identical.
>
> In some cases if the problem is ill conditioned, it helps to add a
small
> amount of gravity in the zero direction of the gravity vector, so
something
> like Gravity={0.001,-9.81,0.001}; may also help solve recruitment
problems
> if this is the root of the problem, it will not change the solution.
>
> If the problems persist, please try to find the parameter set
causing the
> problem and then upload this model to the file section of the
group, then we
> will try to debug it, please also write which version of AnyBody
you are
> using.
>
> About the output, please try to insert an AnyOutPutFile object in
your study
> this will let you export as many variables you like to a text file.
>
> I am uncertain about the installation problem; i will pass this
question on
> the development team.
>
> I hope this helped you solved the problem; otherwise please do not
hesitate
> to write again.
>
> Best regards
>
> SÃ¸ren, AnyBody Support
>
> _____
>
> From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com] On
> Behalf
> Of vtveteran
> Sent: 20 November 2007 13:39
> To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> Subject: [AnyScript] Trouble with multi-dimensional analysis and
home
> installation
>
> We are two students at LTH in Lund, Sweden who are using Anybody
> v3.0.0 to do a biomechanical analysis of a cyclist as part of our
> master thesis project. Unfortunately we have run into a little
> difficulty and hope you can help us with a few points. We have
> adapted the Bike Model 3D program from the repository to run
> parameter and optimisation studies so that we can investigate the
> effects of varying saddle height, saddle position and pedal arm
> width, in the hope of comparing these results to real-world results
> from test subjects.
>
> Our first and greatest problem is that our program is prone to
> unexpected, but reproducible, crashes that simply halt the analysis
> with no error messages given or explanations as to the cause. The
> program gives no compilation errors when loaded, and the crashes
seem
> to be mainly caused by the number of steps we make the program
> investigate. Our most recent crash, for example, involved an n =
> {4,4,4} step analysis, given a certain set starting value and
> variable values for each parameter. A previously run n = {2,2,2}
step
> analysis, with identical start values and variable values, ran
> through to its conclusion with no crashes. We found we could reduce
> the number of crashes by using the simple muscle model as opposed
to
> the 3E model, but they still occur, thus far, every time we attempt
an
> analysis of more than around 40 steps in total.
>
> Our second problem is that we cannot export our results to an
external
> browser. Given our need for a 4-dimensional graph to view our
results
> satisfactorily, the ability to export the results is vitally
> important for us. We realise that it is possible, but are having
> difficulties implemeting the commands found under the Export folder
> in the properties meny of the 2D/3D Chart. Any pointers you could
> give us on how to do this properly would be greatly appreciated!
>
> Thirdly, I have attempted to install Anybody on my home computer,
with
> which we could perform these analyses much faster than on campus,
but
> am having problems with the student license dongle installation. I
am
> running Vista 64-bit, which I realise isn’t ideal for the moment,
but
> I have managed to install all the relevant parts of Anybody itself
> and the License Manager using Windows XP SP2 compatibility mode.
When
> I try to open Anybody, the License window appears, and when I
> press “Dongle Bound License”, a screen pops up telling me to insert
> the dongle. Unfortunately, this happens despite the fact that the
> dongle is already inserted in the USB drive with a constant red
> light, and the HASP License Manager is active (even when in XP SP2
> compatability mode).
>
> We would greatly appreciate any help you can give us! The school PC
> we are trying to run on is a P4 1.8GHz with 512mb RAM with WinXP;
is
> this a factor in the crashes? If so, running it on my home computer
> (Core 2 Duo 2.66GHz, 4gb RAM) is an even more pressing necessity,
> given that Vista isn’t a limiting factor on its own.
>
> Hope you can help us! Thanks in advance. We can of course send you
our
> program if you need it.
>
> Henrik SjÃ¶stedt
> Tobias Lestrell
>
> [Non-text portions of this message have been removed]
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

[Non-text portions of this message have been removed]

system · November 22, 2007, 1:11pm

Thanks again guys,

I thought a saw a reply in my hotmail junk box other than this one,
but unfortunately the genius of Microsoft at their finest conspired
to delete the mail from all my folders when all I did was mark the
mail as safe. If it was anything important, do you think you could
send it again?

Cheers!

Henrik & Tobias

— In anyscript@yahoogroups.com, “AnyBody Support” <support@…>
wrote:
>
> Hi Henrik and Tobias
>
>
>
> I am glad that it works now.
>
>
>
> In the AnyOutPutFile object there are various settings available
please see
> more about this in the reference manual. By default each column is
separated
> by a “,”. So if you in the “Data” menu of Excel choose “import
external
> data” this will let you choose to have a comma as separation sign.
If you
> prefer to have another sign as for the separation this is also
possible.
>
>
>
> Best regards
>
> SÃ¸ren, AnyBody Support
>
>
>
>
>
>
>
> _____
>
> From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com]
On Behalf
> Of vtveteran
> Sent: 22 November 2007 11:01
> To: anyscript@yahoogroups.com
> Subject: [AnyScript] Re: Trouble with multi-dimensional analysis
and home
> installation
>
>
>
> Hi SÃ¸ren and Michael,
>
> Thanks a lot for your advice! Changing the linear algorithm to
> MinMaxSimplex appears to have done the trick, and we’ve run
analyses
> up to 125 steps without issue. We did leave a {10,10,10} analysis
> running overnight that crashed, but this feels like a different
issue
> because instead of the analysis simply halting mid-stride, Windows
> flagged the program as crashed and asked if we wanted to send a
> report, which hasn’t been the case before. We also managed to sort
> out a working output file, though we do wonder if there is any way
> Excel can recognise the different columns on its own.
>
> As concerns the Vista64 issue, I’ll be glad to help out, though I’m
> obviously at uni most of the day, away from my home PC. The issue
I’m
> having at the moment I presume is with the HASP License Manager, as
> opposed to Anybody itself.
>
> Thanks for all the help!
>
> Henrik SjÃ¶stedt
> Tobias Lestrell
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com,
> “AnyBody Support” <support@>
> wrote:
> >
> > Hi Henrik and Tobias,
> >
> >
> >
> > I am writing you regarding the Vista 64 issue. We have only little
> > experience with Vista and in particular Vista 64 yet. We are
> currently
> > preparing to support Vista 32 better in future versions of
AnyBody.
> >
> > I would like to help you as much as I can, but you will have to
> help me too
> > then, i.e., to gain experience into the problem. I would very much
> > appreciate this assistance from you.
> >
> > If you are ready for this I suggest that we take the conversation
> off the
> > group while experimenting.
> >
> >
> >
> > Best regards,
> >
> >
> >
> > Michael
> >
> > ------------------------------
> >
> > Michael Damsgaard, AnyBody Support
> >
> > <http://www.anybodyt <http://www.anybodytech.com> ech.com>
> www.anybodytech.com
> >
> >
> >
> >
> >
> > _____
> >
> > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com]
> On Behalf
> > Of AnyBody Support
> > Sent: Tuesday, November 20, 2007 21:35
> > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> > Subject: RE: [AnyScript] Trouble with multi-dimensional analysis
> and home
> > installation
> >
> >
> >
> > Hi Henrik and Tobias
> >
> > Welcome to the group, it sounds as an interesting problem.
> >
> > From your description it sounds as you may have a recruitment
> problem, but
> > it is difficult to say from the description.
> >
> > To find out if it is recruitment related please try to rerun the
> parameter
> > study but only run the kinematical analysis.
> >
> > If the error is related to recruitment and if the error is
> reproducible my
> > suggestion would be to try to locate which parameter set, in the
> parameters
> > study are causing this error. I think that it may be the
> combination of the
> > parameters rather than the number of steps in the parameter study
> which are
> > causing the crash. So if you try running the parameter study
again
> and then
> > notice for which parameter set it crash? Then subsequently try to
> rerun the
> > model with this set of parameters only and see if the error can be
> > reproduced. If this is that case, then investigate the parameters
> and if
> > possible evaluate if there could be any apparent reason for the
> problem. If
> > this seems not to be the case you could try to change the
> recruitment solver
> > it may possible solve the problem.
> >
> > Please read more about the different solvers in the tutorial
> >
> > http://www.anybodyt <http://www.anybodyt
> <http://www.anybodytech.com/691.0.html> ech.com/691.0.html>
> > ech.com/691.0.html
> >
> > There are two Linear algorithms to choose between MinMaxSimplex
and
> > MinMaxNRSimplex , if you have tried only one of these, please try
> the other
> > one and see if this solves the problem, the results are
identical.
> >
> > In some cases if the problem is ill conditioned, it helps to add
a
> small
> > amount of gravity in the zero direction of the gravity vector, so
> something
> > like Gravity={0.001,-9.81,0.001}; may also help solve recruitment
> problems
> > if this is the root of the problem, it will not change the
solution.
> >
> > If the problems persist, please try to find the parameter set
> causing the
> > problem and then upload this model to the file section of the
> group, then we
> > will try to debug it, please also write which version of AnyBody
> you are
> > using.
> >
> > About the output, please try to insert an AnyOutPutFile object in
> your study
> > this will let you export as many variables you like to a text
file.
> >
> > I am uncertain about the installation problem; i will pass this
> question on
> > the development team.
> >
> > I hope this helped you solved the problem; otherwise please do
not
> hesitate
> > to write again.
> >
> > Best regards
> >
> > SÃ¸ren, AnyBody Support
> >
> > _____
> >
> > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com
> > [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com] On
> > Behalf
> > Of vtveteran
> > Sent: 20 November 2007 13:39
> > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> > Subject: [AnyScript] Trouble with multi-dimensional analysis and
> home
> > installation
> >
> > We are two students at LTH in Lund, Sweden who are using Anybody
> > v3.0.0 to do a biomechanical analysis of a cyclist as part of our
> > master thesis project. Unfortunately we have run into a little
> > difficulty and hope you can help us with a few points. We have
> > adapted the Bike Model 3D program from the repository to run
> > parameter and optimisation studies so that we can investigate the
> > effects of varying saddle height, saddle position and pedal arm
> > width, in the hope of comparing these results to real-world
results
> > from test subjects.
> >
> > Our first and greatest problem is that our program is prone to
> > unexpected, but reproducible, crashes that simply halt the
analysis
> > with no error messages given or explanations as to the cause. The
> > program gives no compilation errors when loaded, and the crashes
> seem
> > to be mainly caused by the number of steps we make the program
> > investigate. Our most recent crash, for example, involved an n =
> > {4,4,4} step analysis, given a certain set starting value and
> > variable values for each parameter. A previously run n = {2,2,2}
> step
> > analysis, with identical start values and variable values, ran
> > through to its conclusion with no crashes. We found we could
reduce
> > the number of crashes by using the simple muscle model as opposed
> to
> > the 3E model, but they still occur, thus far, every time we
attempt
> an
> > analysis of more than around 40 steps in total.
> >
> > Our second problem is that we cannot export our results to an
> external
> > browser. Given our need for a 4-dimensional graph to view our
> results
> > satisfactorily, the ability to export the results is vitally
> > important for us. We realise that it is possible, but are having
> > difficulties implemeting the commands found under the Export
folder
> > in the properties meny of the 2D/3D Chart. Any pointers you could
> > give us on how to do this properly would be greatly appreciated!
> >
> > Thirdly, I have attempted to install Anybody on my home computer,
> with
> > which we could perform these analyses much faster than on campus,
> but
> > am having problems with the student license dongle installation.
I
> am
> > running Vista 64-bit, which I realise isn’t ideal for the moment,
> but
> > I have managed to install all the relevant parts of Anybody
itself
> > and the License Manager using Windows XP SP2 compatibility mode.
> When
> > I try to open Anybody, the License window appears, and when I
> > press “Dongle Bound License”, a screen pops up telling me to
insert
> > the dongle. Unfortunately, this happens despite the fact that the
> > dongle is already inserted in the USB drive with a constant red
> > light, and the HASP License Manager is active (even when in XP
SP2
> > compatability mode).
> >
> > We would greatly appreciate any help you can give us! The school
PC
> > we are trying to run on is a P4 1.8GHz with 512mb RAM with WinXP;
> is
> > this a factor in the crashes? If so, running it on my home
computer
> > (Core 2 Duo 2.66GHz, 4gb RAM) is an even more pressing necessity,
> > given that Vista isn’t a limiting factor on its own.
> >
> > Hope you can help us! Thanks in advance. We can of course send
you
> our
> > program if you need it.
> >
> > Henrik SjÃ¶stedt
> > Tobias Lestrell
> >
> > [Non-text portions of this message have been removed]
> >
> >
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

system · November 26, 2007, 3:54pm

Hi again,

We’ve now got a working model that gives consistent results, although
we aren’t particularly happy with the results themselves. The model
tends to give the lowest values for metabolism when the saddle is
pushed beyond the crank, at values of +0.10m, which we feel shouldn’t
be correct. Your optimisation tutorial gave a nice valley-shaped
curve, whilst the surface our models gives looks more like a
mountain. Could we have a serious problem with the model?

We would like some help in modifying the model to suit different
people. We can change the scale of the lower and upper leg, and the
foot, but we can’t tailor that to specific people unless we know the
length of these appendages at scale = 1.0. We’ve had a poke around
the different classes with Ingrid, our examinator, but we’ve had no
luck so far in finding the specific sizes of the sections.

We would also like some clarification on how the PedalArmWidth is
measured. The description in the Main class says it is the distance
measured between the feet, but purely observational reasoning from
the model would suggest that it’s more likely to be the distance from
the foot to the centre of the crank. The pedal is connected to the
foot via the GroundJoint, but we can’t work out where exactly on the
foot the GroundJoint is. If the feet were placed side-by-side,
touching eachother, would this distance be zero?

The CrankMomentTopDeadCenter value also interests us. Do you have any
information on typical values for this measurement that can be
achieved by top level athletes? We’ve tried doing some googling, but
havn’t found any information on the point. If you have access to any
values with references we would be very grateful if you could share
them with us.

Thanks for all the great help so far, hope you can can help us with
these too!

Henrik SjÃ¶stedt
Tobias Lestrell

— In anyscript@yahoogroups.com, “vtveteran” <henriksjostedt@…>
wrote:
>
> Thanks again guys,
>
> I thought a saw a reply in my hotmail junk box other than this one,
> but unfortunately the genius of Microsoft at their finest conspired
> to delete the mail from all my folders when all I did was mark the
> mail as safe. If it was anything important, do you think you could
> send it again?
>
> Cheers!
>
> Henrik & Tobias
>
> — In anyscript@yahoogroups.com, “AnyBody Support” <support@>
> wrote:
> >
> > Hi Henrik and Tobias
> >
> >
> >
> > I am glad that it works now.
> >
> >
> >
> > In the AnyOutPutFile object there are various settings available
> please see
> > more about this in the reference manual. By default each column
is
> separated
> > by a “,”. So if you in the “Data” menu of Excel choose “import
> external
> > data” this will let you choose to have a comma as separation
sign.
> If you
> > prefer to have another sign as for the separation this is also
> possible.
> >
> >
> >
> > Best regards
> >
> > SÃ¸ren, AnyBody Support
> >
> >
> >
> >
> >
> >
> >
> > ___
> >
> > From: anyscript@yahoogroups.com
[mailto:anyscript@yahoogroups.com]
> On Behalf
> > Of vtveteran
> > Sent: 22 November 2007 11:01
> > To: anyscript@yahoogroups.com
> > Subject: [AnyScript] Re: Trouble with multi-dimensional analysis
> and home
> > installation
> >
> >
> >
> > Hi SÃ¸ren and Michael,
> >
> > Thanks a lot for your advice! Changing the linear algorithm to
> > MinMaxSimplex appears to have done the trick, and we’ve run
> analyses
> > up to 125 steps without issue. We did leave a {10,10,10} analysis
> > running overnight that crashed, but this feels like a different
> issue
> > because instead of the analysis simply halting mid-stride,
Windows
> > flagged the program as crashed and asked if we wanted to send a
> > report, which hasn’t been the case before. We also managed to
sort
> > out a working output file, though we do wonder if there is any
way
> > Excel can recognise the different columns on its own.
> >
> > As concerns the Vista64 issue, I’ll be glad to help out, though
I’m
> > obviously at uni most of the day, away from my home PC. The issue
> I’m
> > having at the moment I presume is with the HASP License Manager,
as
> > opposed to Anybody itself.
> >
> > Thanks for all the help!
> >
> > Henrik SjÃ¶stedt
> > Tobias Lestrell
> >
> > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com,
> > “AnyBody Support” <support@>
> > wrote:
> > >
> > > Hi Henrik and Tobias,
> > >
> > >
> > >
> > > I am writing you regarding the Vista 64 issue. We have only
little
> > > experience with Vista and in particular Vista 64 yet. We are
> > currently
> > > preparing to support Vista 32 better in future versions of
> AnyBody.
> > >
> > > I would like to help you as much as I can, but you will have to
> > help me too
> > > then, i.e., to gain experience into the problem. I would very
much
> > > appreciate this assistance from you.
> > >
> > > If you are ready for this I suggest that we take the
conversation
> > off the
> > > group while experimenting.
> > >
> > >
> > >
> > > Best regards,
> > >
> > >
> > >
> > > Michael
> > >
> > > ------------------------------
> > >
> > > Michael Damsgaard, AnyBody Support
> > >
> > > <http://www.anybodyt <http://www.anybodytech.com> ech.com>
> > www.anybodytech.com
> > >
> > >
> > >
> > >
> > >
> > > _
> > >
> > > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com
> > [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com]
> > On Behalf
> > > Of AnyBody Support
> > > Sent: Tuesday, November 20, 2007 21:35
> > > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com
> > > Subject: RE: [AnyScript] Trouble with multi-dimensional
analysis
> > and home
> > > installation
> > >
> > >
> > >
> > > Hi Henrik and Tobias
> > >
> > > Welcome to the group, it sounds as an interesting problem.
> > >
> > > From your description it sounds as you may have a recruitment
> > problem, but
> > > it is difficult to say from the description.
> > >
> > > To find out if it is recruitment related please try to rerun
the
> > parameter
> > > study but only run the kinematical analysis.
> > >
> > > If the error is related to recruitment and if the error is
> > reproducible my
> > > suggestion would be to try to locate which parameter set, in
the
> > parameters
> > > study are causing this error. I think that it may be the
> > combination of the
> > > parameters rather than the number of steps in the parameter
study
> > which are
> > > causing the crash. So if you try running the parameter study
> again
> > and then
> > > notice for which parameter set it crash? Then subsequently try
to
> > rerun the
> > > model with this set of parameters only and see if the error can
be
> > > reproduced. If this is that case, then investigate the
parameters
> > and if
> > > possible evaluate if there could be any apparent reason for the
> > problem. If
> > > this seems not to be the case you could try to change the
> > recruitment solver
> > > it may possible solve the problem.
> > >
> > > Please read more about the different solvers in the tutorial
> > >
> > > http://www.anybodyt <http://www.anybodyt
> > <http://www.anybodytech.com/691.0.html> ech.com/691.0.html>
> > > ech.com/691.0.html
> > >
> > > There are two Linear algorithms to choose between MinMaxSimplex
> and
> > > MinMaxNRSimplex , if you have tried only one of these, please
try
> > the other
> > > one and see if this solves the problem, the results are
> identical.
> > >
> > > In some cases if the problem is ill conditioned, it helps to
add
> a
> > small
> > > amount of gravity in the zero direction of the gravity vector,
so
> > something
> > > like Gravity={0.001,-9.81,0.001}; may also help solve
recruitment
> > problems
> > > if this is the root of the problem, it will not change the
> solution.
> > >
> > > If the problems persist, please try to find the parameter set
> > causing the
> > > problem and then upload this model to the file section of the
> > group, then we
> > > will try to debug it, please also write which version of
AnyBody
> > you are
> > > using.
> > >
> > > About the output, please try to insert an AnyOutPutFile object
in
> > your study
> > > this will let you export as many variables you like to a text
> file.
> > >
> > > I am uncertain about the installation problem; i will pass this
> > question on
> > > the development team.
> > >
> > > I hope this helped you solved the problem; otherwise please do
> not
> > hesitate
> > > to write again.
> > >
> > > Best regards
> > >
> > > SÃ¸ren, AnyBody Support
> > >
> > > ___
> > >
> > > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> > ps.com
> > > [mailto:anyscript@yahoogrou <mailto:anyscript%
40yahoogroups.com>
> > ps.com] On
> > > Behalf
> > > Of vtveteran
> > > Sent: 20 November 2007 13:39
> > > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com
> > > Subject: [AnyScript] Trouble with multi-dimensional analysis
and
> > home
> > > installation
> > >
> > > We are two students at LTH in Lund, Sweden who are using
Anybody
> > > v3.0.0 to do a biomechanical analysis of a cyclist as part of
our
> > > master thesis project. Unfortunately we have run into a little
> > > difficulty and hope you can help us with a few points. We have
> > > adapted the Bike Model 3D program from the repository to run
> > > parameter and optimisation studies so that we can investigate
the
> > > effects of varying saddle height, saddle position and pedal arm
> > > width, in the hope of comparing these results to real-world
> results
> > > from test subjects.
> > >
> > > Our first and greatest problem is that our program is prone to
> > > unexpected, but reproducible, crashes that simply halt the
> analysis
> > > with no error messages given or explanations as to the cause.
The
> > > program gives no compilation errors when loaded, and the
crashes
> > seem
> > > to be mainly caused by the number of steps we make the program
> > > investigate. Our most recent crash, for example, involved an n

> > > {4,4,4} step analysis, given a certain set starting value and
> > > variable values for each parameter. A previously run n =
{2,2,2}
> > step
> > > analysis, with identical start values and variable values, ran
> > > through to its conclusion with no crashes. We found we could
> reduce
> > > the number of crashes by using the simple muscle model as
opposed
> > to
> > > the 3E model, but they still occur, thus far, every time we
> attempt
> > an
> > > analysis of more than around 40 steps in total.
> > >
> > > Our second problem is that we cannot export our results to an
> > external
> > > browser. Given our need for a 4-dimensional graph to view our
> > results
> > > satisfactorily, the ability to export the results is vitally
> > > important for us. We realise that it is possible, but are
having
> > > difficulties implemeting the commands found under the Export
> folder
> > > in the properties meny of the 2D/3D Chart. Any pointers you
could
> > > give us on how to do this properly would be greatly appreciated!
> > >
> > > Thirdly, I have attempted to install Anybody on my home
computer,
> > with
> > > which we could perform these analyses much faster than on
campus,
> > but
> > > am having problems with the student license dongle
installation.
> I
> > am
> > > running Vista 64-bit, which I realise isn’t ideal for the
moment,
> > but
> > > I have managed to install all the relevant parts of Anybody
> itself
> > > and the License Manager using Windows XP SP2 compatibility
mode.
> > When
> > > I try to open Anybody, the License window appears, and when I
> > > press “Dongle Bound License”, a screen pops up telling me to
> insert
> > > the dongle. Unfortunately, this happens despite the fact that
the
> > > dongle is already inserted in the USB drive with a constant red
> > > light, and the HASP License Manager is active (even when in XP
> SP2
> > > compatability mode).
> > >
> > > We would greatly appreciate any help you can give us! The
school
> PC
> > > we are trying to run on is a P4 1.8GHz with 512mb RAM with
WinXP;
> > is
> > > this a factor in the crashes? If so, running it on my home
> computer
> > > (Core 2 Duo 2.66GHz, 4gb RAM) is an even more pressing
necessity,
> > > given that Vista isn’t a limiting factor on its own.
> > >
> > > Hope you can help us! Thanks in advance. We can of course send
> you
> > our
> > > program if you need it.
> > >
> > > Henrik SjÃ¶stedt
> > > Tobias Lestrell
> > >
> > > [Non-text portions of this message have been removed]
> > >
> > >
> > >
> > >
> > >
> > > [Non-text portions of this message have been removed]
> > >
> >
> >
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
>

system · November 27, 2007, 11:46am

Hi Henrik and Tobias

The difference between the surface plots can be related to the difference in
the models. The surface plot in the tutorial is created with a simple 2d leg
model and with constant strength muscles; the results from this model will
be different from a 3d leg model with three element hill type muscles. It
is difficult to say if there is a problem with your model; the saddle may
also shift in position depending on the load you have applied to the crank.

About scaling the model, please have a look in the Gait3D model this is an
example of a model which is scaled to fit a certain sized person. If you
look in the AnyMan.any file you will find a section with segment lengths,
here you simply insert the segment lengths you would like to have. In the
main file you then need to specify a scaling law you will get the best
results by using the ScaleLengthMassFat.any law. In the gait3d.amun.any file
you can see how to include the scalinglaw (although here it is another law
which has been used).

About the goundnode I do not exactly remember its location but you can
always create an AnyDrawRefFrame to make a node location visibleâ€¦.

So in this case you could add one line to the PedalFootJoint
(JointsandDrivers.any)

AnyRevoluteJoint PedalFootJoint = {

 AnyRefNode &ref1= ....BikeModel.Crank.RPedal;

AnyRefNode &ref2 = …ModelFolder.Right.Leg.Seg.Foot.GroundJoint;

Ref2={AnyDrawRefFrame Drw={};}; //this lines adds a visible coordinate
system in the ref2 node, and since the ref2 node points at the groundjoint
at the foot it will display it there … you could also have inserted it
directly where the GroundJointNode was created

 #include

“…/…/…/Body/AAUHuman/ToolBox/JointReactions/DrawJointReactions.any”

 AnyFolder &DrawRef=Main.DrawSettings;

};

You are right about the pedalarmwidth this is measured from crank center to
the Cramk.Rpedal node for example so it gives half the distance between the
peldals.

This can be seen here in the definition of the crank

AnySeg Crank = {

 Mass=0.5;

 Jii = {0.001,0.001,0.001};

 AnyRefNode CrankJoint = { sRel = {0,0,0}; };

 AnyRefNode RPedal = { sRel =

{0,…BikeParameters.PedalArmLength,…BikeParameters.PedalArmWidth}; };

 AnyRefNode LPedal = { sRel =

{0,-…BikeParameters.PedalArmLength,-…BikeParameters.PedalArmWidth}; };

 AnyDrawSurf CrankSurf ={

   FileName = "crank";

   ScaleXYZ={0.9,0.9,0.9};

   RGB=Main.DrawSettings.Colors.AnyBodyGreen;

 };

};//Crank

I do not have not recall having references with max values for
CrankTopDeadCenter, I am sorry.

Please ask again if you have further questions.

Best regards

SÃ¸ren

From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of vtveteran
Sent: 26 November 2007 15:55
To: anyscript@yahoogroups.com
Subject: [AnyScript] Re: Trouble with multi-dimensional analysis and home
installation

Hi again,

We’ve now got a working model that gives consistent results, although
we aren’t particularly happy with the results themselves. The model
tends to give the lowest values for metabolism when the saddle is
pushed beyond the crank, at values of +0.10m, which we feel shouldn’t
be correct. Your optimisation tutorial gave a nice valley-shaped
curve, whilst the surface our models gives looks more like a
mountain. Could we have a serious problem with the model?

We would like some help in modifying the model to suit different
people. We can change the scale of the lower and upper leg, and the
foot, but we can’t tailor that to specific people unless we know the
length of these appendages at scale = 1.0. We’ve had a poke around
the different classes with Ingrid, our examinator, but we’ve had no
luck so far in finding the specific sizes of the sections.

We would also like some clarification on how the PedalArmWidth is
measured. The description in the Main class says it is the distance
measured between the feet, but purely observational reasoning from
the model would suggest that it’s more likely to be the distance from
the foot to the centre of the crank. The pedal is connected to the
foot via the GroundJoint, but we can’t work out where exactly on the
foot the GroundJoint is. If the feet were placed side-by-side,
touching eachother, would this distance be zero?

The CrankMomentTopDeadCenter value also interests us. Do you have any
information on typical values for this measurement that can be
achieved by top level athletes? We’ve tried doing some googling, but
havn’t found any information on the point. If you have access to any
values with references we would be very grateful if you could share
them with us.

Thanks for all the great help so far, hope you can can help us with
these too!

Henrik SjÃ¶stedt
Tobias Lestrell

— In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
“vtveteran” <henriksjostedt@…>
wrote:
>
> Thanks again guys,
>
> I thought a saw a reply in my hotmail junk box other than this one,
> but unfortunately the genius of Microsoft at their finest conspired
> to delete the mail from all my folders when all I did was mark the
> mail as safe. If it was anything important, do you think you could
> send it again?
>
> Cheers!
>
> Henrik & Tobias
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
“AnyBody Support” <support@>
> wrote:
> >
> > Hi Henrik and Tobias
> >
> >
> >
> > I am glad that it works now.
> >
> >
> >
> > In the AnyOutPutFile object there are various settings available
> please see
> > more about this in the reference manual. By default each column
is
> separated
> > by a “,”. So if you in the “Data” menu of Excel choose “import
> external
> > data” this will let you choose to have a comma as separation
sign.
> If you
> > prefer to have another sign as for the separation this is also
> possible.
> >
> >
> >
> > Best regards
> >
> > SÃ¸ren, AnyBody Support
> >
> >
> >
> >
> >
> >
> >
> > ___
> >
> > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
[mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com]
> On Behalf
> > Of vtveteran
> > Sent: 22 November 2007 11:01
> > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> > Subject: [AnyScript] Re: Trouble with multi-dimensional analysis
> and home
> > installation
> >
> >
> >
> > Hi SÃ¸ren and Michael,
> >
> > Thanks a lot for your advice! Changing the linear algorithm to
> > MinMaxSimplex appears to have done the trick, and we’ve run
> analyses
> > up to 125 steps without issue. We did leave a {10,10,10} analysis
> > running overnight that crashed, but this feels like a different
> issue
> > because instead of the analysis simply halting mid-stride,
Windows
> > flagged the program as crashed and asked if we wanted to send a
> > report, which hasn’t been the case before. We also managed to
sort
> > out a working output file, though we do wonder if there is any
way
> > Excel can recognise the different columns on its own.
> >
> > As concerns the Vista64 issue, I’ll be glad to help out, though
I’m
> > obviously at uni most of the day, away from my home PC. The issue
> I’m
> > having at the moment I presume is with the HASP License Manager,
as
> > opposed to Anybody itself.
> >
> > Thanks for all the help!
> >
> > Henrik SjÃ¶stedt
> > Tobias Lestrell
> >
> > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com,
> > “AnyBody Support” <support@>
> > wrote:
> > >
> > > Hi Henrik and Tobias,
> > >
> > >
> > >
> > > I am writing you regarding the Vista 64 issue. We have only
little
> > > experience with Vista and in particular Vista 64 yet. We are
> > currently
> > > preparing to support Vista 32 better in future versions of
> AnyBody.
> > >
> > > I would like to help you as much as I can, but you will have to
> > help me too
> > > then, i.e., to gain experience into the problem. I would very
much
> > > appreciate this assistance from you.
> > >
> > > If you are ready for this I suggest that we take the
conversation
> > off the
> > > group while experimenting.
> > >
> > >
> > >
> > > Best regards,
> > >
> > >
> > >
> > > Michael
> > >
> > > ------------------------------
> > >
> > > Michael Damsgaard, AnyBody Support
> > >
> > > <http://www.anybodyt <http://www.anybodyt <http://www.anybodytech.com>
ech.com> ech.com>
> > www.anybodytech.com
> > >
> > >
> > >
> > >
> > >
> > > _
> > >
> > > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com
> > [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com]
> > On Behalf
> > > Of AnyBody Support
> > > Sent: Tuesday, November 20, 2007 21:35
> > > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com
> > > Subject: RE: [AnyScript] Trouble with multi-dimensional
analysis
> > and home
> > > installation
> > >
> > >
> > >
> > > Hi Henrik and Tobias
> > >
> > > Welcome to the group, it sounds as an interesting problem.
> > >
> > > From your description it sounds as you may have a recruitment
> > problem, but
> > > it is difficult to say from the description.
> > >
> > > To find out if it is recruitment related please try to rerun
the
> > parameter
> > > study but only run the kinematical analysis.
> > >
> > > If the error is related to recruitment and if the error is
> > reproducible my
> > > suggestion would be to try to locate which parameter set, in
the
> > parameters
> > > study are causing this error. I think that it may be the
> > combination of the
> > > parameters rather than the number of steps in the parameter
study
> > which are
> > > causing the crash. So if you try running the parameter study
> again
> > and then
> > > notice for which parameter set it crash? Then subsequently try
to
> > rerun the
> > > model with this set of parameters only and see if the error can
be
> > > reproduced. If this is that case, then investigate the
parameters
> > and if
> > > possible evaluate if there could be any apparent reason for the
> > problem. If
> > > this seems not to be the case you could try to change the
> > recruitment solver
> > > it may possible solve the problem.
> > >
> > > Please read more about the different solvers in the tutorial
> > >
> > > http://www.anybodyt <http://www.anybodyt
> > <http://www.anybodyt <http://www.anybodytech.com/691.0.html>
ech.com/691.0.html> ech.com/691.0.html>
> > > ech.com/691.0.html
> > >
> > > There are two Linear algorithms to choose between MinMaxSimplex
> and
> > > MinMaxNRSimplex , if you have tried only one of these, please
try
> > the other
> > > one and see if this solves the problem, the results are
> identical.
> > >
> > > In some cases if the problem is ill conditioned, it helps to
add
> a
> > small
> > > amount of gravity in the zero direction of the gravity vector,
so
> > something
> > > like Gravity={0.001,-9.81,0.001}; may also help solve
recruitment
> > problems
> > > if this is the root of the problem, it will not change the
> solution.
> > >
> > > If the problems persist, please try to find the parameter set
> > causing the
> > > problem and then upload this model to the file section of the
> > group, then we
> > > will try to debug it, please also write which version of
AnyBody
> > you are
> > > using.
> > >
> > > About the output, please try to insert an AnyOutPutFile object
in
> > your study
> > > this will let you export as many variables you like to a text
> file.
> > >
> > > I am uncertain about the installation problem; i will pass this
> > question on
> > > the development team.
> > >
> > > I hope this helped you solved the problem; otherwise please do
> not
> > hesitate
> > > to write again.
> > >
> > > Best regards
> > >
> > > SÃ¸ren, AnyBody Support
> > >
> > > ___
> > >
> > > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> > ps.com
> > > [mailto:anyscript@yahoogrou <mailto:anyscript%
40yahoogroups.com>
> > ps.com] On
> > > Behalf
> > > Of vtveteran
> > > Sent: 20 November 2007 13:39
> > > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com
> > > Subject: [AnyScript] Trouble with multi-dimensional analysis
and
> > home
> > > installation
> > >
> > > We are two students at LTH in Lund, Sweden who are using
Anybody
> > > v3.0.0 to do a biomechanical analysis of a cyclist as part of
our
> > > master thesis project. Unfortunately we have run into a little
> > > difficulty and hope you can help us with a few points. We have
> > > adapted the Bike Model 3D program from the repository to run
> > > parameter and optimisation studies so that we can investigate
the
> > > effects of varying saddle height, saddle position and pedal arm
> > > width, in the hope of comparing these results to real-world
> results
> > > from test subjects.
> > >
> > > Our first and greatest problem is that our program is prone to
> > > unexpected, but reproducible, crashes that simply halt the
> analysis
> > > with no error messages given or explanations as to the cause.
The
> > > program gives no compilation errors when loaded, and the
crashes
> > seem
> > > to be mainly caused by the number of steps we make the program
> > > investigate. Our most recent crash, for example, involved an n

> > > {4,4,4} step analysis, given a certain set starting value and
> > > variable values for each parameter. A previously run n =
{2,2,2}
> > step
> > > analysis, with identical start values and variable values, ran
> > > through to its conclusion with no crashes. We found we could
> reduce
> > > the number of crashes by using the simple muscle model as
opposed
> > to
> > > the 3E model, but they still occur, thus far, every time we
> attempt
> > an
> > > analysis of more than around 40 steps in total.
> > >
> > > Our second problem is that we cannot export our results to an
> > external
> > > browser. Given our need for a 4-dimensional graph to view our
> > results
> > > satisfactorily, the ability to export the results is vitally
> > > important for us. We realise that it is possible, but are
having
> > > difficulties implemeting the commands found under the Export
> folder
> > > in the properties meny of the 2D/3D Chart. Any pointers you
could
> > > give us on how to do this properly would be greatly appreciated!
> > >
> > > Thirdly, I have attempted to install Anybody on my home
computer,
> > with
> > > which we could perform these analyses much faster than on
campus,
> > but
> > > am having problems with the student license dongle
installation.
> I
> > am
> > > running Vista 64-bit, which I realise isn’t ideal for the
moment,
> > but
> > > I have managed to install all the relevant parts of Anybody
> itself
> > > and the License Manager using Windows XP SP2 compatibility
mode.
> > When
> > > I try to open Anybody, the License window appears, and when I
> > > press “Dongle Bound License”, a screen pops up telling me to
> insert
> > > the dongle. Unfortunately, this happens despite the fact that
the
> > > dongle is already inserted in the USB drive with a constant red
> > > light, and the HASP License Manager is active (even when in XP
> SP2
> > > compatability mode).
> > >
> > > We would greatly appreciate any help you can give us! The
school
> PC
> > > we are trying to run on is a P4 1.8GHz with 512mb RAM with
WinXP;
> > is
> > > this a factor in the crashes? If so, running it on my home
> computer
> > > (Core 2 Duo 2.66GHz, 4gb RAM) is an even more pressing
necessity,
> > > given that Vista isn’t a limiting factor on its own.
> > >
> > > Hope you can help us! Thanks in advance. We can of course send
> you
> > our
> > > program if you need it.
> > >
> > > Henrik SjÃ¶stedt
> > > Tobias Lestrell
> > >
> > > [Non-text portions of this message have been removed]
> > >
> > >
> > >
> > >
> > >
> > > [Non-text portions of this message have been removed]
> > >
> >
> >
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
>

[Non-text portions of this message have been removed]

system · November 27, 2007, 8:41pm

Hello Henrik and Tobias,

It just occurred to me that it is necessary to phase shift the crank
torque profile and possibly also the ankle angle variation when you
move the seat back and forth. I don’t think this is implemented in
the 3-D bicycle model.

Best regards,
John

— In anyscript@yahoogroups.com, “AnyBody Support” <support@…>
wrote:
>
> Hi Henrik and Tobias
>
>
>
> The difference between the surface plots can be related to the
difference in
> the models. The surface plot in the tutorial is created with a
simple 2d leg
> model and with constant strength muscles; the results from this
model will
> be different from a 3d leg model with three element hill type
muscles. It
> is difficult to say if there is a problem with your model; the
saddle may
> also shift in position depending on the load you have applied to
the crank.
>
>
>
> About scaling the model, please have a look in the Gait3D model
this is an
> example of a model which is scaled to fit a certain sized person.
If you
> look in the AnyMan.any file you will find a section with segment
lengths,
> here you simply insert the segment lengths you would like to have.
In the
> main file you then need to specify a scaling law you will get the
best
> results by using the ScaleLengthMassFat.any law. In the
gait3d.amun.any file
> you can see how to include the scalinglaw (although here it is
another law
> which has been used).
>
>
>
> About the goundnode I do not exactly remember its location but you
can
> always create an AnyDrawRefFrame to make a node location visibleâ€¦.
>
>
>
> So in this case you could add one line to the PedalFootJoint
> (JointsandDrivers.any)
>
>
>
>
>
> AnyRevoluteJoint PedalFootJoint = {
>
> AnyRefNode &ref1= …BikeModel.Crank.RPedal;
>
>
>
> AnyRefNode &ref2 = …ModelFolder.Right.Leg.Seg.Foot.GroundJoint;
>
> Ref2={AnyDrawRefFrame Drw={};}; //this lines adds a visible
coordinate
> system in the ref2 node, and since the ref2 node points at the
groundjoint
> at the foot it will display it there … you could also have
inserted it
> directly where the GroundJointNode was created
>
> #include
> “…/…/…/Body/AAUHuman/ToolBox/JointReactions/DrawJointReactions.an
y”
>
>
>
> AnyFolder &DrawRef=Main.DrawSettings;
>
>
>
> };
>
>
>
>
>
> You are right about the pedalarmwidth this is measured from crank
center to
> the Cramk.Rpedal node for example so it gives half the distance
between the
> peldals.
>
> This can be seen here in the definition of the crank
>
>
>
> AnySeg Crank = {
>
> Mass=0.5;
>
> Jii = {0.001,0.001,0.001};
>
> AnyRefNode CrankJoint = { sRel = {0,0,0}; };
>
> AnyRefNode RPedal = { sRel =
>
{0,…BikeParameters.PedalArmLength,…BikeParameters.PedalArmWidth}
; };
>
>
> AnyRefNode LPedal = { sRel =
> {0,-…BikeParameters.PedalArmLength,-
…BikeParameters.PedalArmWidth}; };
>
>
> AnyDrawSurf CrankSurf ={
>
> FileName = “crank”;
>
> ScaleXYZ={0.9,0.9,0.9};
>
> RGB=Main.DrawSettings.Colors.AnyBodyGreen;
>
> };
>
> };//Crank
>
>
>
> I do not have not recall having references with max values for
> CrankTopDeadCenter, I am sorry.
>
>
>
> Please ask again if you have further questions.
>
>
>
> Best regards
>
> SÃ¸ren
>
> _____
>
> From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com]
On Behalf
> Of vtveteran
> Sent: 26 November 2007 15:55
> To: anyscript@yahoogroups.com
> Subject: [AnyScript] Re: Trouble with multi-dimensional analysis
and home
> installation
>
>
>
> Hi again,
>
> We’ve now got a working model that gives consistent results,
although
> we aren’t particularly happy with the results themselves. The model
> tends to give the lowest values for metabolism when the saddle is
> pushed beyond the crank, at values of +0.10m, which we feel
shouldn’t
> be correct. Your optimisation tutorial gave a nice valley-shaped
> curve, whilst the surface our models gives looks more like a
> mountain. Could we have a serious problem with the model?
>
> We would like some help in modifying the model to suit different
> people. We can change the scale of the lower and upper leg, and the
> foot, but we can’t tailor that to specific people unless we know
the
> length of these appendages at scale = 1.0. We’ve had a poke around
> the different classes with Ingrid, our examinator, but we’ve had no
> luck so far in finding the specific sizes of the sections.
>
> We would also like some clarification on how the PedalArmWidth is
> measured. The description in the Main class says it is the distance
> measured between the feet, but purely observational reasoning from
> the model would suggest that it’s more likely to be the distance
from
> the foot to the centre of the crank. The pedal is connected to the
> foot via the GroundJoint, but we can’t work out where exactly on
the
> foot the GroundJoint is. If the feet were placed side-by-side,
> touching eachother, would this distance be zero?
>
> The CrankMomentTopDeadCenter value also interests us. Do you have
any
> information on typical values for this measurement that can be
> achieved by top level athletes? We’ve tried doing some googling,
but
> havn’t found any information on the point. If you have access to
any
> values with references we would be very grateful if you could share
> them with us.
>
> Thanks for all the great help so far, hope you can can help us with
> these too!
>
> Henrik SjÃ¶stedt
> Tobias Lestrell
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com,
> “vtveteran” <henriksjostedt@>
> wrote:
> >
> > Thanks again guys,
> >
> > I thought a saw a reply in my hotmail junk box other than this
one,
> > but unfortunately the genius of Microsoft at their finest
conspired
> > to delete the mail from all my folders when all I did was mark
the
> > mail as safe. If it was anything important, do you think you
could
> > send it again?
> >
> > Cheers!
> >
> > Henrik & Tobias
> >
> > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com,
> “AnyBody Support” <support@>
> > wrote:
> > >
> > > Hi Henrik and Tobias
> > >
> > >
> > >
> > > I am glad that it works now.
> > >
> > >
> > >
> > > In the AnyOutPutFile object there are various settings
available
> > please see
> > > more about this in the reference manual. By default each column
> is
> > separated
> > > by a “,”. So if you in the “Data” menu of Excel choose “import
> > external
> > > data” this will let you choose to have a comma as separation
> sign.
> > If you
> > > prefer to have another sign as for the separation this is also
> > possible.
> > >
> > >
> > >
> > > Best regards
> > >
> > > SÃ¸ren, AnyBody Support
> > >
> > >
> > >
> > >
> > >
> > >
> > >
> > > _____
> > >
> > > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com]
> > On Behalf
> > > Of vtveteran
> > > Sent: 22 November 2007 11:01
> > > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com
> > > Subject: [AnyScript] Re: Trouble with multi-dimensional
analysis
> > and home
> > > installation
> > >
> > >
> > >
> > > Hi SÃ¸ren and Michael,
> > >
> > > Thanks a lot for your advice! Changing the linear algorithm to
> > > MinMaxSimplex appears to have done the trick, and we’ve run
> > analyses
> > > up to 125 steps without issue. We did leave a {10,10,10}
analysis
> > > running overnight that crashed, but this feels like a different
> > issue
> > > because instead of the analysis simply halting mid-stride,
> Windows
> > > flagged the program as crashed and asked if we wanted to send a
> > > report, which hasn’t been the case before. We also managed to
> sort
> > > out a working output file, though we do wonder if there is any
> way
> > > Excel can recognise the different columns on its own.
> > >
> > > As concerns the Vista64 issue, I’ll be glad to help out, though
> I’m
> > > obviously at uni most of the day, away from my home PC. The
issue
> > I’m
> > > having at the moment I presume is with the HASP License
Manager,
> as
> > > opposed to Anybody itself.
> > >
> > > Thanks for all the help!
> > >
> > > Henrik SjÃ¶stedt
> > > Tobias Lestrell
> > >
> > > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> > ps.com,
> > > “AnyBody Support” <support@>
> > > wrote:
> > > >
> > > > Hi Henrik and Tobias,
> > > >
> > > >
> > > >
> > > > I am writing you regarding the Vista 64 issue. We have only
> little
> > > > experience with Vista and in particular Vista 64 yet. We are
> > > currently
> > > > preparing to support Vista 32 better in future versions of
> > AnyBody.
> > > >
> > > > I would like to help you as much as I can, but you will have
to
> > > help me too
> > > > then, i.e., to gain experience into the problem. I would very
> much
> > > > appreciate this assistance from you.
> > > >
> > > > If you are ready for this I suggest that we take the
> conversation
> > > off the
> > > > group while experimenting.
> > > >
> > > >
> > > >
> > > > Best regards,
> > > >
> > > >
> > > >
> > > > Michael
> > > >
> > > > ------------------------------
> > > >
> > > > Michael Damsgaard, AnyBody Support
> > > >
> > > > <http://www.anybodyt <http://www.anybodyt
<http://www.anybodytech.com>
> ech.com> ech.com>
> > > www.anybodytech.com
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > _____
> > > >
> > > > From: anyscript@yahoogrou <mailto:anyscript%
40yahoogroups.com>
> > ps.com
> > > [mailto:anyscript@yahoogrou <mailto:anyscript%
40yahoogroups.com>
> > ps.com]
> > > On Behalf
> > > > Of AnyBody Support
> > > > Sent: Tuesday, November 20, 2007 21:35
> > > > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> > ps.com
> > > > Subject: RE: [AnyScript] Trouble with multi-dimensional
> analysis
> > > and home
> > > > installation
> > > >
> > > >
> > > >
> > > > Hi Henrik and Tobias
> > > >
> > > > Welcome to the group, it sounds as an interesting problem.
> > > >
> > > > From your description it sounds as you may have a recruitment
> > > problem, but
> > > > it is difficult to say from the description.
> > > >
> > > > To find out if it is recruitment related please try to rerun
> the
> > > parameter
> > > > study but only run the kinematical analysis.
> > > >
> > > > If the error is related to recruitment and if the error is
> > > reproducible my
> > > > suggestion would be to try to locate which parameter set, in
> the
> > > parameters
> > > > study are causing this error. I think that it may be the
> > > combination of the
> > > > parameters rather than the number of steps in the parameter
> study
> > > which are
> > > > causing the crash. So if you try running the parameter study
> > again
> > > and then
> > > > notice for which parameter set it crash? Then subsequently
try
> to
> > > rerun the
> > > > model with this set of parameters only and see if the error
can
> be
> > > > reproduced. If this is that case, then investigate the
> parameters
> > > and if
> > > > possible evaluate if there could be any apparent reason for
the
> > > problem. If
> > > > this seems not to be the case you could try to change the
> > > recruitment solver
> > > > it may possible solve the problem.
> > > >
> > > > Please read more about the different solvers in the tutorial
> > > >
> > > > http://www.anybodyt <http://www.anybodyt
> > > <http://www.anybodyt <http://www.anybodytech.com/691.0.html>
> ech.com/691.0.html> ech.com/691.0.html>
> > > > ech.com/691.0.html
> > > >
> > > > There are two Linear algorithms to choose between
MinMaxSimplex
> > and
> > > > MinMaxNRSimplex , if you have tried only one of these, please
> try
> > > the other
> > > > one and see if this solves the problem, the results are
> > identical.
> > > >
> > > > In some cases if the problem is ill conditioned, it helps to
> add
> > a
> > > small
> > > > amount of gravity in the zero direction of the gravity
vector,
> so
> > > something
> > > > like Gravity={0.001,-9.81,0.001}; may also help solve
> recruitment
> > > problems
> > > > if this is the root of the problem, it will not change the
> > solution.
> > > >
> > > > If the problems persist, please try to find the parameter set
> > > causing the
> > > > problem and then upload this model to the file section of the
> > > group, then we
> > > > will try to debug it, please also write which version of
> AnyBody
> > > you are
> > > > using.
> > > >
> > > > About the output, please try to insert an AnyOutPutFile
object
> in
> > > your study
> > > > this will let you export as many variables you like to a text
> > file.
> > > >
> > > > I am uncertain about the installation problem; i will pass
this
> > > question on
> > > > the development team.
> > > >
> > > > I hope this helped you solved the problem; otherwise please
do
> > not
> > > hesitate
> > > > to write again.
> > > >
> > > > Best regards
> > > >
> > > > SÃ¸ren, AnyBody Support
> > > >
> > > > _____
> > > >
> > > > From: anyscript@yahoogrou <mailto:anyscript%
40yahoogroups.com>
> > > ps.com
> > > > [mailto:anyscript@yahoogrou <mailto:anyscript%
> 40yahoogroups.com>
> > > ps.com] On
> > > > Behalf
> > > > Of vtveteran
> > > > Sent: 20 November 2007 13:39
> > > > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> > ps.com
> > > > Subject: [AnyScript] Trouble with multi-dimensional analysis
> and
> > > home
> > > > installation
> > > >
> > > > We are two students at LTH in Lund, Sweden who are using
> Anybody
> > > > v3.0.0 to do a biomechanical analysis of a cyclist as part of
> our
> > > > master thesis project. Unfortunately we have run into a
little
> > > > difficulty and hope you can help us with a few points. We
have
> > > > adapted the Bike Model 3D program from the repository to run
> > > > parameter and optimisation studies so that we can investigate
> the
> > > > effects of varying saddle height, saddle position and pedal
arm
> > > > width, in the hope of comparing these results to real-world
> > results
> > > > from test subjects.
> > > >
> > > > Our first and greatest problem is that our program is prone
to
> > > > unexpected, but reproducible, crashes that simply halt the
> > analysis
> > > > with no error messages given or explanations as to the cause.
> The
> > > > program gives no compilation errors when loaded, and the
> crashes
> > > seem
> > > > to be mainly caused by the number of steps we make the
program
> > > > investigate. Our most recent crash, for example, involved an
n
> =
> > > > {4,4,4} step analysis, given a certain set starting value and
> > > > variable values for each parameter. A previously run n =
> {2,2,2}
> > > step
> > > > analysis, with identical start values and variable values,
ran
> > > > through to its conclusion with no crashes. We found we could
> > reduce
> > > > the number of crashes by using the simple muscle model as
> opposed
> > > to
> > > > the 3E model, but they still occur, thus far, every time we
> > attempt
> > > an
> > > > analysis of more than around 40 steps in total.
> > > >
> > > > Our second problem is that we cannot export our results to an
> > > external
> > > > browser. Given our need for a 4-dimensional graph to view our
> > > results
> > > > satisfactorily, the ability to export the results is vitally
> > > > important for us. We realise that it is possible, but are
> having
> > > > difficulties implemeting the commands found under the Export
> > folder
> > > > in the properties meny of the 2D/3D Chart. Any pointers you
> could
> > > > give us on how to do this properly would be greatly
appreciated!
> > > >
> > > > Thirdly, I have attempted to install Anybody on my home
> computer,
> > > with
> > > > which we could perform these analyses much faster than on
> campus,
> > > but
> > > > am having problems with the student license dongle
> installation.
> > I
> > > am
> > > > running Vista 64-bit, which I realise isn’t ideal for the
> moment,
> > > but
> > > > I have managed to install all the relevant parts of Anybody
> > itself
> > > > and the License Manager using Windows XP SP2 compatibility
> mode.
> > > When
> > > > I try to open Anybody, the License window appears, and when I
> > > > press “Dongle Bound License”, a screen pops up telling me to
> > insert
> > > > the dongle. Unfortunately, this happens despite the fact that
> the
> > > > dongle is already inserted in the USB drive with a constant
red
> > > > light, and the HASP License Manager is active (even when in
XP
> > SP2
> > > > compatability mode).
> > > >
> > > > We would greatly appreciate any help you can give us! The
> school
> > PC
> > > > we are trying to run on is a P4 1.8GHz with 512mb RAM with
> WinXP;
> > > is
> > > > this a factor in the crashes? If so, running it on my home
> > computer
> > > > (Core 2 Duo 2.66GHz, 4gb RAM) is an even more pressing
> necessity,
> > > > given that Vista isn’t a limiting factor on its own.
> > > >
> > > > Hope you can help us! Thanks in advance. We can of course
send
> > you
> > > our
> > > > program if you need it.
> > > >
> > > > Henrik SjÃ¶stedt
> > > > Tobias Lestrell
> > > >
> > > > [Non-text portions of this message have been removed]
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > [Non-text portions of this message have been removed]
> > > >
> > >
> > >
> > >
> > >
> > >
> > > [Non-text portions of this message have been removed]
> > >
> >
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

system · November 29, 2007, 10:43am

Hej John!

Nu har vi fler problem:

Vi har fÃ¶rstÃ¥tt att b1,b2 ger B i Fourierdrivern. De Ã¤r satta till
konstanter. Ã„r det de som vi skall Ã¤ndra?
Finns det nÃ¥gon ledtrÃ¥d i nÃ¥got som ni har skrivit om hur vi skall
bestÃ¤mma dessa konstanter? Hur funkar det?

Sen undrar vi ocksÃ¥ hur foten sitter fast i pedalen?
Tror du det Ã¤r smart att lÃ¤gga in en extra nod pÃ¥ foten som vi
fÃ¶rbinder med pedalen?
Det ser lite konstigt ut pÃ¥ modellen bilden. Det ser ut som han
trampar med hÃ¥lfoten!

HÃ¤lsningar
Ingrid
med flera

— In anyscript@yahoogroups.com, “AnyBody Support” <support@…>
wrote:
>
> Hello Henrik and Tobias,
>
> It just occurred to me that it is necessary to phase shift the
crank
> torque profile and possibly also the ankle angle variation when you
> move the seat back and forth. I don’t think this is implemented in
> the 3-D bicycle model.
>
> Best regards,
> John
>
>
> — In anyscript@yahoogroups.com, “AnyBody Support” <support@>
> wrote:
> >
> > Hi Henrik and Tobias
> >
> >
> >
> > The difference between the surface plots can be related to the
> difference in
> > the models. The surface plot in the tutorial is created with a
> simple 2d leg
> > model and with constant strength muscles; the results from this
> model will
> > be different from a 3d leg model with three element hill type
> muscles. It
> > is difficult to say if there is a problem with your model; the
> saddle may
> > also shift in position depending on the load you have applied to
> the crank.
> >
> >
> >
> > About scaling the model, please have a look in the Gait3D model
> this is an
> > example of a model which is scaled to fit a certain sized person.
> If you
> > look in the AnyMan.any file you will find a section with segment
> lengths,
> > here you simply insert the segment lengths you would like to
have.
> In the
> > main file you then need to specify a scaling law you will get the
> best
> > results by using the ScaleLengthMassFat.any law. In the
> gait3d.amun.any file
> > you can see how to include the scalinglaw (although here it is
> another law
> > which has been used).
> >
> >
> >
> > About the goundnode I do not exactly remember its location but
you
> can
> > always create an AnyDrawRefFrame to make a node location visibleâ€¦.
> >
> >
> >
> > So in this case you could add one line to the PedalFootJoint
> > (JointsandDrivers.any)
> >
> >
> >
> >
> >
> > AnyRevoluteJoint PedalFootJoint = {
> >
> > AnyRefNode &ref1= …BikeModel.Crank.RPedal;
> >
> >
> >
> > AnyRefNode &ref2
= …ModelFolder.Right.Leg.Seg.Foot.GroundJoint;
> >
> > Ref2={AnyDrawRefFrame Drw={};}; //this lines adds a visible
> coordinate
> > system in the ref2 node, and since the ref2 node points at the
> groundjoint
> > at the foot it will display it there … you could also have
> inserted it
> > directly where the GroundJointNode was created
> >
> > #include
>
> “…/…/…/Body/AAUHuman/ToolBox/JointReactions/DrawJointReactions.an
> y”
> >
> >
> >
> > AnyFolder &DrawRef=Main.DrawSettings;
> >
> >
> >
> > };
> >
> >
> >
> >
> >
> > You are right about the pedalarmwidth this is measured from crank
> center to
> > the Cramk.Rpedal node for example so it gives half the distance
> between the
> > peldals.
> >
> > This can be seen here in the definition of the crank
> >
> >
> >
> > AnySeg Crank = {
> >
> > Mass=0.5;
> >
> > Jii = {0.001,0.001,0.001};
> >
> > AnyRefNode CrankJoint = { sRel = {0,0,0}; };
> >
> > AnyRefNode RPedal = { sRel =
> >
>
{0,…BikeParameters.PedalArmLength,…BikeParameters.PedalArmWidth}
> ; };
> >
> >
> > AnyRefNode LPedal = { sRel =
> > {0,-…BikeParameters.PedalArmLength,-
> …BikeParameters.PedalArmWidth}; };
> >
> >
> > AnyDrawSurf CrankSurf ={
> >
> > FileName = “crank”;
> >
> > ScaleXYZ={0.9,0.9,0.9};
> >
> > RGB=Main.DrawSettings.Colors.AnyBodyGreen;
> >
> > };
> >
> > };//Crank
> >
> >
> >
> > I do not have not recall having references with max values for
> > CrankTopDeadCenter, I am sorry.
> >
> >
> >
> > Please ask again if you have further questions.
> >
> >
> >
> > Best regards
> >
> > SÃ¸ren
> >
> > _____
> >
> > From: anyscript@yahoogroups.com
[mailto:anyscript@yahoogroups.com]
> On Behalf
> > Of vtveteran
> > Sent: 26 November 2007 15:55
> > To: anyscript@yahoogroups.com
> > Subject: [AnyScript] Re: Trouble with multi-dimensional analysis
> and home
> > installation
> >
> >
> >
> > Hi again,
> >
> > We’ve now got a working model that gives consistent results,
> although
> > we aren’t particularly happy with the results themselves. The
model
> > tends to give the lowest values for metabolism when the saddle is
> > pushed beyond the crank, at values of +0.10m, which we feel
> shouldn’t
> > be correct. Your optimisation tutorial gave a nice valley-shaped
> > curve, whilst the surface our models gives looks more like a
> > mountain. Could we have a serious problem with the model?
> >
> > We would like some help in modifying the model to suit different
> > people. We can change the scale of the lower and upper leg, and
the
> > foot, but we can’t tailor that to specific people unless we know
> the
> > length of these appendages at scale = 1.0. We’ve had a poke
around
> > the different classes with Ingrid, our examinator, but we’ve had
no
> > luck so far in finding the specific sizes of the sections.
> >
> > We would also like some clarification on how the PedalArmWidth is
> > measured. The description in the Main class says it is the
distance
> > measured between the feet, but purely observational reasoning
from
> > the model would suggest that it’s more likely to be the distance
> from
> > the foot to the centre of the crank. The pedal is connected to
the
> > foot via the GroundJoint, but we can’t work out where exactly on
> the
> > foot the GroundJoint is. If the feet were placed side-by-side,
> > touching eachother, would this distance be zero?
> >
> > The CrankMomentTopDeadCenter value also interests us. Do you have
> any
> > information on typical values for this measurement that can be
> > achieved by top level athletes? We’ve tried doing some googling,
> but
> > havn’t found any information on the point. If you have access to
> any
> > values with references we would be very grateful if you could
share
> > them with us.
> >
> > Thanks for all the great help so far, hope you can can help us
with
> > these too!
> >
> > Henrik SjÃ¶stedt
> > Tobias Lestrell
> >
> > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com,
> > “vtveteran” <henriksjostedt@>
> > wrote:
> > >
> > > Thanks again guys,
> > >
> > > I thought a saw a reply in my hotmail junk box other than this
> one,
> > > but unfortunately the genius of Microsoft at their finest
> conspired
> > > to delete the mail from all my folders when all I did was mark
> the
> > > mail as safe. If it was anything important, do you think you
> could
> > > send it again?
> > >
> > > Cheers!
> > >
> > > Henrik & Tobias
> > >
> > > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com,
> > “AnyBody Support” <support@>
> > > wrote:
> > > >
> > > > Hi Henrik and Tobias
> > > >
> > > >
> > > >
> > > > I am glad that it works now.
> > > >
> > > >
> > > >
> > > > In the AnyOutPutFile object there are various settings
> available
> > > please see
> > > > more about this in the reference manual. By default each
column
> > is
> > > separated
> > > > by a “,”. So if you in the “Data” menu of Excel
choose “import
> > > external
> > > > data” this will let you choose to have a comma as separation
> > sign.
> > > If you
> > > > prefer to have another sign as for the separation this is
also
> > > possible.
> > > >
> > > >
> > > >
> > > > Best regards
> > > >
> > > > SÃ¸ren, AnyBody Support
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > _____
> > > >
> > > > From: anyscript@yahoogrou <mailto:anyscript%
40yahoogroups.com>
> ps.com
> > [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com]
> > > On Behalf
> > > > Of vtveteran
> > > > Sent: 22 November 2007 11:01
> > > > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com
> > > > Subject: [AnyScript] Re: Trouble with multi-dimensional
> analysis
> > > and home
> > > > installation
> > > >
> > > >
> > > >
> > > > Hi SÃ¸ren and Michael,
> > > >
> > > > Thanks a lot for your advice! Changing the linear algorithm
to
> > > > MinMaxSimplex appears to have done the trick, and we’ve run
> > > analyses
> > > > up to 125 steps without issue. We did leave a {10,10,10}
> analysis
> > > > running overnight that crashed, but this feels like a
different
> > > issue
> > > > because instead of the analysis simply halting mid-stride,
> > Windows
> > > > flagged the program as crashed and asked if we wanted to send
a
> > > > report, which hasn’t been the case before. We also managed to
> > sort
> > > > out a working output file, though we do wonder if there is
any
> > way
> > > > Excel can recognise the different columns on its own.
> > > >
> > > > As concerns the Vista64 issue, I’ll be glad to help out,
though
> > I’m
> > > > obviously at uni most of the day, away from my home PC. The
> issue
> > > I’m
> > > > having at the moment I presume is with the HASP License
> Manager,
> > as
> > > > opposed to Anybody itself.
> > > >
> > > > Thanks for all the help!
> > > >
> > > > Henrik SjÃ¶stedt
> > > > Tobias Lestrell
> > > >
> > > > — In anyscript@yahoogrou <mailto:anyscript%
40yahoogroups.com>
> > > ps.com,
> > > > “AnyBody Support” <support@>
> > > > wrote:
> > > > >
> > > > > Hi Henrik and Tobias,
> > > > >
> > > > >
> > > > >
> > > > > I am writing you regarding the Vista 64 issue. We have only
> > little
> > > > > experience with Vista and in particular Vista 64 yet. We
are
> > > > currently
> > > > > preparing to support Vista 32 better in future versions of
> > > AnyBody.
> > > > >
> > > > > I would like to help you as much as I can, but you will
have
> to
> > > > help me too
> > > > > then, i.e., to gain experience into the problem. I would
very
> > much
> > > > > appreciate this assistance from you.
> > > > >
> > > > > If you are ready for this I suggest that we take the
> > conversation
> > > > off the
> > > > > group while experimenting.
> > > > >
> > > > >
> > > > >
> > > > > Best regards,
> > > > >
> > > > >
> > > > >
> > > > > Michael
> > > > >
> > > > > ------------------------------
> > > > >
> > > > > Michael Damsgaard, AnyBody Support
> > > > >
> > > > > <http://www.anybodyt <http://www.anybodyt
> <http://www.anybodytech.com>
> > ech.com> ech.com>
> > > > www.anybodytech.com
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > _____
> > > > >
> > > > > From: anyscript@yahoogrou <mailto:anyscript%
> 40yahoogroups.com>
> > > ps.com
> > > > [mailto:anyscript@yahoogrou <mailto:anyscript%
> 40yahoogroups.com>
> > > ps.com]
> > > > On Behalf
> > > > > Of AnyBody Support
> > > > > Sent: Tuesday, November 20, 2007 21:35
> > > > > To: anyscript@yahoogrou <mailto:anyscript%
40yahoogroups.com>
> > > ps.com
> > > > > Subject: RE: [AnyScript] Trouble with multi-dimensional
> > analysis
> > > > and home
> > > > > installation
> > > > >
> > > > >
> > > > >
> > > > > Hi Henrik and Tobias
> > > > >
> > > > > Welcome to the group, it sounds as an interesting problem.
> > > > >
> > > > > From your description it sounds as you may have a
recruitment
> > > > problem, but
> > > > > it is difficult to say from the description.
> > > > >
> > > > > To find out if it is recruitment related please try to
rerun
> > the
> > > > parameter
> > > > > study but only run the kinematical analysis.
> > > > >
> > > > > If the error is related to recruitment and if the error is
> > > > reproducible my
> > > > > suggestion would be to try to locate which parameter set,
in
> > the
> > > > parameters
> > > > > study are causing this error. I think that it may be the
> > > > combination of the
> > > > > parameters rather than the number of steps in the parameter
> > study
> > > > which are
> > > > > causing the crash. So if you try running the parameter
study
> > > again
> > > > and then
> > > > > notice for which parameter set it crash? Then subsequently
> try
> > to
> > > > rerun the
> > > > > model with this set of parameters only and see if the error
> can
> > be
> > > > > reproduced. If this is that case, then investigate the
> > parameters
> > > > and if
> > > > > possible evaluate if there could be any apparent reason for
> the
> > > > problem. If
> > > > > this seems not to be the case you could try to change the
> > > > recruitment solver
> > > > > it may possible solve the problem.
> > > > >
> > > > > Please read more about the different solvers in the
tutorial
> > > > >
> > > > > http://www.anybodyt <http://www.anybodyt
> > > > <http://www.anybodyt <http://www.anybodytech.com/691.0.html>
> > ech.com/691.0.html> ech.com/691.0.html>
> > > > > ech.com/691.0.html
> > > > >
> > > > > There are two Linear algorithms to choose between
> MinMaxSimplex
> > > and
> > > > > MinMaxNRSimplex , if you have tried only one of these,
please
> > try
> > > > the other
> > > > > one and see if this solves the problem, the results are
> > > identical.
> > > > >
> > > > > In some cases if the problem is ill conditioned, it helps
to
> > add
> > > a
> > > > small
> > > > > amount of gravity in the zero direction of the gravity
> vector,
> > so
> > > > something
> > > > > like Gravity={0.001,-9.81,0.001}; may also help solve
> > recruitment
> > > > problems
> > > > > if this is the root of the problem, it will not change the
> > > solution.
> > > > >
> > > > > If the problems persist, please try to find the parameter
set
> > > > causing the
> > > > > problem and then upload this model to the file section of
the
> > > > group, then we
> > > > > will try to debug it, please also write which version of
> > AnyBody
> > > > you are
> > > > > using.
> > > > >
> > > > > About the output, please try to insert an AnyOutPutFile
> object
> > in
> > > > your study
> > > > > this will let you export as many variables you like to a
text
> > > file.
> > > > >
> > > > > I am uncertain about the installation problem; i will pass
> this
> > > > question on
> > > > > the development team.
> > > > >
> > > > > I hope this helped you solved the problem; otherwise please
> do
> > > not
> > > > hesitate
> > > > > to write again.
> > > > >
> > > > > Best regards
> > > > >
> > > > > SÃ¸ren, AnyBody Support
> > > > >
> > > > > _____
> > > > >
> > > > > From: anyscript@yahoogrou <mailto:anyscript%
> 40yahoogroups.com>
> > > > ps.com
> > > > > [mailto:anyscript@yahoogrou <mailto:anyscript%
> > 40yahoogroups.com>
> > > > ps.com] On
> > > > > Behalf
> > > > > Of vtveteran
> > > > > Sent: 20 November 2007 13:39
> > > > > To: anyscript@yahoogrou <mailto:anyscript%
40yahoogroups.com>
> > > ps.com
> > > > > Subject: [AnyScript] Trouble with multi-dimensional
analysis
> > and
> > > > home
> > > > > installation
> > > > >
> > > > > We are two students at LTH in Lund, Sweden who are using
> > Anybody
> > > > > v3.0.0 to do a biomechanical analysis of a cyclist as part
of
> > our
> > > > > master thesis project. Unfortunately we have run into a
> little
> > > > > difficulty and hope you can help us with a few points. We
> have
> > > > > adapted the Bike Model 3D program from the repository to
run
> > > > > parameter and optimisation studies so that we can
investigate
> > the
> > > > > effects of varying saddle height, saddle position and pedal
> arm
> > > > > width, in the hope of comparing these results to real-world
> > > results
> > > > > from test subjects.
> > > > >
> > > > > Our first and greatest problem is that our program is prone
> to
> > > > > unexpected, but reproducible, crashes that simply halt the
> > > analysis
> > > > > with no error messages given or explanations as to the
cause.
> > The
> > > > > program gives no compilation errors when loaded, and the
> > crashes
> > > > seem
> > > > > to be mainly caused by the number of steps we make the
> program
> > > > > investigate. Our most recent crash, for example, involved
an
> n
> > =
> > > > > {4,4,4} step analysis, given a certain set starting value
and
> > > > > variable values for each parameter. A previously run n =
> > {2,2,2}
> > > > step
> > > > > analysis, with identical start values and variable values,
> ran
> > > > > through to its conclusion with no crashes. We found we
could
> > > reduce
> > > > > the number of crashes by using the simple muscle model as
> > opposed
> > > > to
> > > > > the 3E model, but they still occur, thus far, every time we
> > > attempt
> > > > an
> > > > > analysis of more than around 40 steps in total.
> > > > >
> > > > > Our second problem is that we cannot export our results to
an
> > > > external
> > > > > browser. Given our need for a 4-dimensional graph to view
our
> > > > results
> > > > > satisfactorily, the ability to export the results is
vitally
> > > > > important for us. We realise that it is possible, but are
> > having
> > > > > difficulties implemeting the commands found under the
Export
> > > folder
> > > > > in the properties meny of the 2D/3D Chart. Any pointers you
> > could
> > > > > give us on how to do this properly would be greatly
> appreciated!
> > > > >
> > > > > Thirdly, I have attempted to install Anybody on my home
> > computer,
> > > > with
> > > > > which we could perform these analyses much faster than on
> > campus,
> > > > but
> > > > > am having problems with the student license dongle
> > installation.
> > > I
> > > > am
> > > > > running Vista 64-bit, which I realise isn’t ideal for the
> > moment,
> > > > but
> > > > > I have managed to install all the relevant parts of Anybody
> > > itself
> > > > > and the License Manager using Windows XP SP2 compatibility
> > mode.
> > > > When
> > > > > I try to open Anybody, the License window appears, and when
I
> > > > > press “Dongle Bound License”, a screen pops up telling me
to
> > > insert
> > > > > the dongle. Unfortunately, this happens despite the fact
that
> > the
> > > > > dongle is already inserted in the USB drive with a constant
> red
> > > > > light, and the HASP License Manager is active (even when in
> XP
> > > SP2
> > > > > compatability mode).
> > > > >
> > > > > We would greatly appreciate any help you can give us! The
> > school
> > > PC
> > > > > we are trying to run on is a P4 1.8GHz with 512mb RAM with
> > WinXP;
> > > > is
> > > > > this a factor in the crashes? If so, running it on my home
> > > computer
> > > > > (Core 2 Duo 2.66GHz, 4gb RAM) is an even more pressing
> > necessity,
> > > > > given that Vista isn’t a limiting factor on its own.
> > > > >
> > > > > Hope you can help us! Thanks in advance. We can of course
> send
> > > you
> > > > our
> > > > > program if you need it.
> > > > >
> > > > > Henrik SjÃ¶stedt
> > > > > Tobias Lestrell
> > > > >
> > > > > [Non-text portions of this message have been removed]
> > > > >
> > > > >
> > > > >
> > > > >
> > > > >
> > > > > [Non-text portions of this message have been removed]
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > >
> > > > [Non-text portions of this message have been removed]
> > > >
> > >
> >
> >
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
>

system · November 29, 2007, 2:55pm

Wow, the group’s first posting in Swedish! Luckily we AnyBuddies are
very multilingual, but I’ll answer in English anyway for the benefit
of the other users.

The construction of the Foruier driver function is pretty much
structured like you will find it in any text book on mathematical
analysis. However, rather than me explaining a whole lot, please have
a look at the “Parameter studies and optimization” tutorial here:
http://www.anybodytech.com/654.0.html.

This tutorial uses a bicycle model where the joint torque is phase
shifted just like you want. You can download the model from a link
inside the tutorial.

The phase shift is accomplished in the main file by the following
line:
AnyVar PhaseShift = atan(SaddlePos/SaddleHeight);

The variable PhaseShift is subsequently used in the definition of the
crank torque.

Best regards,
John

— In anyscript@yahoogroups.com, “vtveteran” <henriksjostedt@…>
wrote:
>
> Hej John!
>
> Nu har vi fler problem:
>
> Vi har fÃ¶rstÃ¥tt att b1,b2 ger B i Fourierdrivern. De Ã¤r satta till
> konstanter. Ã„r det de som vi skall Ã¤ndra?
> Finns det nÃ¥gon ledtrÃ¥d i nÃ¥got som ni har skrivit om hur vi skall
> bestÃ¤mma dessa konstanter? Hur funkar det?
>
> Sen undrar vi ocksÃ¥ hur foten sitter fast i pedalen?
> Tror du det Ã¤r smart att lÃ¤gga in en extra nod pÃ¥ foten som vi
> fÃ¶rbinder med pedalen?
> Det ser lite konstigt ut pÃ¥ modellen bilden. Det ser ut som han
> trampar med hÃ¥lfoten!
>
> HÃ¤lsningar
> Ingrid
> med flera
>

system · November 29, 2007, 3:58pm

Hi John!

Now we feel stupid. We trawled all through the JointsAndDrivers
file of the OptimStudy looking for differences to the one in Bike3D,
and only finding slashed-out text adding an extra dimension to the A
and B variables in the Fourier driver. It never occurred to us that
it could be something so simple as a variable in the main file.
Anyway, I’m curious as to how this is implemented. It feels odd that
we merely state these values (the phase shift and all the wiggles) as
variables in the main file and then they magically perform their
function. Don’t we have to implement them into the method somehow?

We also tried to change the foot position on the pedal by switching
from GroundJoint to ToeJoint in the definition of both the foot-joint
connection and the line of rotation of the foot-pedal. This has
however had no effect whatsoever on the image that appears in the
model window; the forces on the pedals (we presume the green arrows
that vary throughout the cycle are force avatars) still eminate from
the centre of the foot, where we don’t want them!!

We would also like to know more about the mechanism rotating the
pedal when the foot is not on the down stroke. The force arrows, if
that is what they are, show no evidence of the leg pulling the pedal
up on the backstroke, as professional cyclists do. Is this feature
implemented into the model? Given the foot placement on the pedal and
the low default mechanical output of 170W, we presume that the model
is more of bike-as-transport cyclist, rather than a professional
racer, which is what we are looking to model. We presume that a pro
is capable of putting out a much larger CrankMomentTopDeadCentre than
1.5, but we can’t seem to find any literature on that anywhere.

Thanks for all your help so far, it is very much appreciated!

Greetings

Henrik SjÃ¶stedt
Tobias Lestrell

— In anyscript@yahoogroups.com, “AnyBody Support” <support@…>
wrote:
>
> Wow, the group’s first posting in Swedish! Luckily we AnyBuddies
are
> very multilingual, but I’ll answer in English anyway for the
benefit
> of the other users.
>
> The construction of the Foruier driver function is pretty much
> structured like you will find it in any text book on mathematical
> analysis. However, rather than me explaining a whole lot, please
have
> a look at the “Parameter studies and optimization” tutorial here:
> http://www.anybodytech.com/654.0.html.
>
> This tutorial uses a bicycle model where the joint torque is phase
> shifted just like you want. You can download the model from a link
> inside the tutorial.
>
> The phase shift is accomplished in the main file by the following
> line:
> AnyVar PhaseShift = atan(SaddlePos/SaddleHeight);
>
> The variable PhaseShift is subsequently used in the definition of
the
> crank torque.
>
> Best regards,
> John
>
> — In anyscript@yahoogroups.com, “vtveteran” <henriksjostedt@>
> wrote:
> >
> > Hej John!
> >
> > Nu har vi fler problem:
> >
> > Vi har fÃ¶rstÃ¥tt att b1,b2 ger B i Fourierdrivern. De Ã¤r satta
till
> > konstanter. Ã„r det de som vi skall Ã¤ndra?
> > Finns det nÃ¥gon ledtrÃ¥d i nÃ¥got som ni har skrivit om hur vi
skall
> > bestÃ¤mma dessa konstanter? Hur funkar det?
> >
> > Sen undrar vi ocksÃ¥ hur foten sitter fast i pedalen?
> > Tror du det Ã¤r smart att lÃ¤gga in en extra nod pÃ¥ foten som vi
> > fÃ¶rbinder med pedalen?
> > Det ser lite konstigt ut pÃ¥ modellen bilden. Det ser ut som han
> > trampar med hÃ¥lfoten!
> >
> > HÃ¤lsningar
> > Ingrid
> > med flera
> >
>

system · November 29, 2007, 10:56pm

Hej,

Just a quick comment here.
One would indeed expect that professional cyclists always will have a
positive tangential pedal force output when using pedals like SPD or
LOOK during steady state cycling on flat road. But that is actually not
the case. Also well trained cyclist have a negative tangential pedal
force in the backstroke. I think the following article sheds some light
on this issue:

Effect of pedaling technique on mechanical effectiveness and efficiency
in cyclists.
Korff T, Romer LM, Mayhew I, Martin JC
<javascript:AL_get(this, ‘jour’, ‘Med Sci Sports Exerc.’);>Med Sci
Sports Exerc. 2007 Jun;39(6):991-5.

If you don’t have access to this article I can send you the PDF.

Greetings, Mark

>
> We would also like to know more about the mechanism rotating the
> pedal when the foot is not on the down stroke. The force arrows, if
> that is what they are, show no evidence of the leg pulling the pedal
> up on the backstroke, as professional cyclists do.
>
>
>

–
Mark de Zee, Ph.D., Post doc.
Aalborg University, Department of Health Science and Technology
Center for Sensory-Motor Interaction (SMI)
Fredrik Bajers Vej 7 D-3, 9220 Aalborg, Denmark
Phone: +45-96358818; Mobile: +45-29809367
E-mail: mdz@hst.aau.dk

system · November 30, 2007, 9:50am

Hi Mark,

We’d love to get a look at that article! We feel that any article on
the subject can be useful to us at this stage.

Thanks in advance

Henrik
Tobias

— In anyscript@yahoogroups.com, Mark de Zee <mdz@…> wrote:
>
> Hej,
>
> Just a quick comment here.
> One would indeed expect that professional cyclists always will have
a
> positive tangential pedal force output when using pedals like SPD
or
> LOOK during steady state cycling on flat road. But that is actually
not
> the case. Also well trained cyclist have a negative tangential
pedal
> force in the backstroke. I think the following article sheds some
light
> on this issue:
>
> Effect of pedaling technique on mechanical effectiveness and
efficiency
> in cyclists.
> Korff T, Romer LM, Mayhew I, Martin JC
> <javascript:AL_get(this, ‘jour’, ‘Med Sci Sports Exerc.’);>Med Sci
> Sports Exerc. 2007 Jun;39(6):991-5.
>
> If you don’t have access to this article I can send you the PDF.
>
> Greetings, Mark
>
> >
> > We would also like to know more about the mechanism rotating the
> > pedal when the foot is not on the down stroke. The force arrows,
if
> > that is what they are, show no evidence of the leg pulling the
pedal
> > up on the backstroke, as professional cyclists do.
> >
> >
> >
>
>
> –
> Mark de Zee, Ph.D., Post doc.
> Aalborg University, Department of Health Science and Technology
> Center for Sensory-Motor Interaction (SMI)
> Fredrik Bajers Vej 7 D-3, 9220 Aalborg, Denmark
> Phone: +45-96358818; Mobile: +45-29809367
> E-mail: mdz@…
>

system · November 30, 2007, 10:17am

Hello Henrik and Tobias!

maybe some of those articles might be of interest to you too:
(some are in german unfortunately, but who knows)
GONZALES, H.; HULL, M.L.: Bivariate Optimization of Pedaling Rate and
Crank Arm
Length in Cycling. J. Biomechanics 21 (1988), 839-849

GONZALES, H.; HULL, M.L : Multivariable Optimization of Cycling
Biomechanics. J.
Biomechanics 22 (1989), 1151-1161

HENKE, T.: Zur biomechanischen Validierung von Komponenten der
Fahrtechnik im
StraÃŸenradsport. Dissertation, Potsdam 1993

KAUTZ, S.A; HULL, M.L.: A theoretical basis for interpreting the force
applied to the pedal
in cycling. Journal of biomechanics 2 (1993), 155-165

MCLEAN, B.; LAFORTUNE, M.: Improving pedalling technique with
“real-time” biomechanical
feedback. Excel 5 (1988), 15-18

NEPTUNE, R.R.; HULL, M.L.: Accuracy Assessment of Methods for
Determining Hip
Movement in Seated Cycling. Journal of biomechanics 28 (1995), 423-437

SANDERSON, D.J.: Training with biofeedback. Real-time feedback improves
pedalling
style. Bike tech 6 (1987), 10-13

SIEBERTZ, K.: Biomechanische Belastungsanalysen unter BerÃ¼cksichtigung
der Leichtbauweise
des Bewegungsapparates. DÃ¼sseldorf VDI-Verlag (1994) (in Druck)

SIEBERTZ, K.; HEMEKE, E.; BAUMANN, W.; GRÃ–NEMEYER, D.; WENTZ, K.:
Kernspinresonanztomographie zur Datenerfassung in der Biomechanik.
Biomedizinische
Technik 38 (1993), 81-86

SUHR, S.: Zur Beurteilung der Fahrtechnik im StraÃŸenradsport â€“ eine
empirische Untersuchung
an unterschiedlich leistungsfÃ¤higen Fahrern zur Erstellung eines
differentiellen
Analysebogens. Diplomarbeit, Bochum 1992

and:
Baron, R.; Bachl, N.; Petschnig, R.; Tschan, H.; Smekal, G.; Pokan,
R.: Measurement of Maximal Power Output in Isokinetic and
Non-Isokinetic Cycling. A Comparison of Two Methods. International
Journal of Sports Medicine, New-York - Stuttgart, 1999, pp. 438-443.

Ericson, M.O.; Nisell, R.: Efficiency of Pedal Forces During Ergometer
Cycling. International Journal of Sports Medicine, 2/1988,
pp.118-122.

[16] Hull, M.L.; Kautz, S.; Beard, A.: An angular velocity Profile on
Cycling
derived from mechanical Energy Analysis. Journal of Biomechanics,
7/1991, pp. 577-586.

Kautz, S.A.; Hull, M.L.: A theoretical Basis fÃ¼r interpreting the Force
applied to the Pedal in Cycling. Journal of Biomechanics, GB,
1993/2, pp. 155-165.

Burke, E., ed.: High-Tech Cycling, Human Kinetics, 2003.

Neptune, R.R.; Herzog, W.: The association between negative
muscle work and pedaling rate. Journal of Biomechanics, 1999, pp.
1021-1026.

Wilson, D.: Bicycling Science., MIT Press, London, 2004.

regards,
stefan

vtveteran schrieb:
> Hi Mark,
>
> We’d love to get a look at that article! We feel that any article on
> the subject can be useful to us at this stage.
>
> Thanks in advance
>
> Henrik
> Tobias
>
> — In anyscript@yahoogroups.com, Mark de Zee <mdz@…> wrote:
>
>> Hej,
>>
>> Just a quick comment here.
>> One would indeed expect that professional cyclists always will have
>>
> a
>
>> positive tangential pedal force output when using pedals like SPD
>>
> or
>
>> LOOK during steady state cycling on flat road. But that is actually
>>
> not
>
>> the case. Also well trained cyclist have a negative tangential
>>
> pedal
>
>> force in the backstroke. I think the following article sheds some
>>
> light
>
>> on this issue:
>>
>> Effect of pedaling technique on mechanical effectiveness and
>>
> efficiency
>
>> in cyclists.
>> Korff T, Romer LM, Mayhew I, Martin JC
>> <javascript:AL_get(this, ‘jour’, ‘Med Sci Sports Exerc.’);>Med Sci
>> Sports Exerc. 2007 Jun;39(6):991-5.
>>
>> If you don’t have access to this article I can send you the PDF.
>>
>> Greetings, Mark
>>
>>
>>> We would also like to know more about the mechanism rotating the
>>> pedal when the foot is not on the down stroke. The force arrows,
>>>
> if
>
>>> that is what they are, show no evidence of the leg pulling the
>>>
> pedal
>
>>> up on the backstroke, as professional cyclists do.
>>>
>>>
>>>
>>>
>> –
>> Mark de Zee, Ph.D., Post doc.
>> Aalborg University, Department of Health Science and Technology
>> Center for Sensory-Motor Interaction (SMI)
>> Fredrik Bajers Vej 7 D-3, 9220 Aalborg, Denmark
>> Phone: +45-96358818; Mobile: +45-29809367
>> E-mail: mdz@…
>>
>>
>
>
>
>
> Users of the AnyBody Modeling System help each other create biomechanical
models in the AnyScript language.
> Yahoo! Groups Links
>
>
>
>
>

system · November 30, 2007, 1:27pm

Hi Henrik and Tobias

I am a little uncertain about the first question, but i will try to explain
how it works.

The crank torque is defined in the file “BikeFrameAndWheel.any” by the
following lines:

//This is the definiton of the crank moment

AnyForce CrankMoment ={

 AnyJoint &Jnt=.CrankJoint;

F={…BikeParameters.CrankMomentOffset-…BikeParameters.CrankMomentAmpcos(
4pi*t/…BikeParameters.T+ …BikeParameters.CrankMomentPhase)};

};

As you can see the torque is related to the parameters in the main file so
these settings control the torque.

In the same file you will also find the definition of the crank motion

AnyKinEqSimpleDriver CrankRotation = {

 AnyRevoluteJoint &Jnt = .CrankJoint;

 DriverPos = {0};

 DriverVel = {-(...BikeParameters.Cadence/60)*pi*2};

 Reaction.Type = {Off}; //This has to be zero otherwise the crank driver

will carry the applied load

};

Again you can see that it directly refers to the parameters of the main
file.

When looking at the ankle motion this is as you know controlled by a fourier
driver in the JointsAndDrivers.any file please see

AnyKinEqFourierDriver PedalRRotationFourier ={

 AnyKinMeasureOrg &Jnt = .FootRMovement;

 Type = CosSin;

 AnyVar a0 = -0.3797-0.5*pi;

 AnyVar a1 = -0.0205;

 AnyVar b1 = 0.3393;

 AnyVar a2 = 0.004906;

 AnyVar b2 = 0.03654;

 AnyVar w = 7.681;

 Freq = ..BikeParameters.Cadence/60;

 A = { {a0+0.5*pi,a1,a2} };

 B = {{0,b1,b2} };

 Reaction.Type = {Off};

};

This motion is only linked to the parameters of the main file in terms of
the cadency.

Concerning the foot pedal joint I tried to dumb the sRel values of the two
nodes ToeJoint and GroundJoint you mention, this is the result:

-------------------------------Objects dump--------------------------

Main.Model.HumanModel.Right.Leg.Seg.Foot.ToeJoint.sRel = {-0, -0.0671,
0.025};

-------------------------------Objects dump--------------------------

Main.Model.HumanModel.Right.Leg.Seg.Foot.GroundJoint.sRel = {0, 0, 0};

There is a big difference in these nodes so shifting the node in the joint
should be visible once the kinematic has been solved (not a load time!)

I tried to do this change and it worked as intended, the joint moved to the
front of the foot. Below is the modified joints for the right side

AnyRevoluteJoint PedalFootJoint = {

 AnyRefNode &ref1= ....BikeModel.Crank.RPedal;

 AnyRefNode &ref2 = ...ModelFolder.Right.Leg.Seg.Foot.ToeJoint;

 #include

“…/…/…/BRep/Aalborg/ToolBox/JointReactions/DrawJointReactions.any”

 AnyFolder &DrawRef=Main.DrawSettings;

};

// This measure is for controlling the rotation.

AnyKinRotational PedalJointRot = {

 AnyRefFrame &Ground = ....BikeModel.BikeFrame;

 AnyRefNode &Foot = ...ModelFolder.Right.Leg.Seg.Foot.ToeJoint;

 Type=RotAxesAngles;

};

About the backstroke: If it will be of any advantage in terms of lowering
the activities of the muscles the model will pull the leg upward at the
backstroke. There is nothing in the model that prevents this from happing;
the foot is attached to the pedal through a revolute joint which can also
carry such a pulling load. So if the model does not show this it is likely
to be related to the loading conditions and motion.

I hope this answered the questions otherwise please ask again.

Best regards

SÃ¸ren, anybody Support

From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of vtveteran
Sent: 29 November 2007 15:59
To: anyscript@yahoogroups.com
Subject: [AnyScript] Re: Trouble with multi-dimensional analysis and home
installation

Hi John!

Now we feel stupid. We trawled all through the JointsAndDrivers
file of the OptimStudy looking for differences to the one in Bike3D,
and only finding slashed-out text adding an extra dimension to the A
and B variables in the Fourier driver. It never occurred to us that
it could be something so simple as a variable in the main file.
Anyway, I’m curious as to how this is implemented. It feels odd that
we merely state these values (the phase shift and all the wiggles) as
variables in the main file and then they magically perform their
function. Don’t we have to implement them into the method somehow?

We also tried to change the foot position on the pedal by switching
from GroundJoint to ToeJoint in the definition of both the foot-joint
connection and the line of rotation of the foot-pedal. This has
however had no effect whatsoever on the image that appears in the
model window; the forces on the pedals (we presume the green arrows
that vary throughout the cycle are force avatars) still eminate from
the centre of the foot, where we don’t want them!!

We would also like to know more about the mechanism rotating the
pedal when the foot is not on the down stroke. The force arrows, if
that is what they are, show no evidence of the leg pulling the pedal
up on the backstroke, as professional cyclists do. Is this feature
implemented into the model? Given the foot placement on the pedal and
the low default mechanical output of 170W, we presume that the model
is more of bike-as-transport cyclist, rather than a professional
racer, which is what we are looking to model. We presume that a pro
is capable of putting out a much larger CrankMomentTopDeadCentre than
1.5, but we can’t seem to find any literature on that anywhere.

Thanks for all your help so far, it is very much appreciated!

Greetings

Henrik SjÃ¶stedt
Tobias Lestrell

— In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
“AnyBody Support” <support@…>
wrote:
>
> Wow, the group’s first posting in Swedish! Luckily we AnyBuddies
are
> very multilingual, but I’ll answer in English anyway for the
benefit
> of the other users.
>
> The construction of the Foruier driver function is pretty much
> structured like you will find it in any text book on mathematical
> analysis. However, rather than me explaining a whole lot, please
have
> a look at the “Parameter studies and optimization” tutorial here:
> http://www.anybodyt <http://www.anybodytech.com/654.0.html.>
ech.com/654.0.html.
>
> This tutorial uses a bicycle model where the joint torque is phase
> shifted just like you want. You can download the model from a link
> inside the tutorial.
>
> The phase shift is accomplished in the main file by the following
> line:
> AnyVar PhaseShift = atan(SaddlePos/SaddleHeight);
>
> The variable PhaseShift is subsequently used in the definition of
the
> crank torque.
>
> Best regards,
> John
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
“vtveteran” <henriksjostedt@>
> wrote:
> >
> > Hej John!
> >
> > Nu har vi fler problem:
> >
> > Vi har fÃ¶rstÃ¥tt att b1,b2 ger B i Fourierdrivern. De Ã¤r satta
till
> > konstanter. Ã„r det de som vi skall Ã¤ndra?
> > Finns det nÃ¥gon ledtrÃ¥d i nÃ¥got som ni har skrivit om hur vi
skall
> > bestÃ¤mma dessa konstanter? Hur funkar det?
> >
> > Sen undrar vi ocksÃ¥ hur foten sitter fast i pedalen?
> > Tror du det Ã¤r smart att lÃ¤gga in en extra nod pÃ¥ foten som vi
> > fÃ¶rbinder med pedalen?
> > Det ser lite konstigt ut pÃ¥ modellen bilden. Det ser ut som han
> > trampar med hÃ¥lfoten!
> >
> > HÃ¤lsningar
> > Ingrid
> > med flera
> >
>

[Non-text portions of this message have been removed]

system · November 30, 2007, 3:37pm

Hi SÃ¶ren!

You guys really are fantastic when it comes to helping us, thanks a
lot!

Sadly, we still can’t get the pedal connection to shift on the foot.
The lines in our Leg.Seg folder relating to the problem are:

AnyRefNode Toetip = {
sRel = .Scale({-0.012,-0.126,0.023});
};

AnyRefNode GroundJoint = {
sRel = .Scale({0.0, 0.0, …Sign*0.0});
ARel = {{0, 1, 0}, {-1, 0, 0}, {0, 0, 1}};
};

AnyRefNode HeelJoint = {
sRel = .Scale({0.009, 0.115, …Sign*0.00});
};

AnyRefNode ToeJoint = {
sRel = .Scale({-0.0, -0.0671, …Sign*0.025});
ARel = {{0, 1, 0}, {-1, 0, 0}, {0, 0, 1}};
};

and the lines for the right hand side in JointsAndDrivers:

AnyFolder Right={
// This is the joint connecting the foot to the pedal. It simply
// specifies that the linear distances between the two must be zero

/AnyKinLinear PedalJointLin = {
AnyRefNode &PedalR= …BikeModel.Crank.RPedal;
AnyRefNode &Foot = …ModelFolder.Right.Leg.Seg.Foot.GroundJoint;
};/

AnyRevoluteJoint PedalFootJoint = {
AnyRefNode &ref1= …BikeModel.Crank.RPedal;
AnyRefNode &ref2 = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;

#include “…/…/…/BRep/Aalborg/ToolBox/JointReactions/DrawJointReacti
ons.any”

 AnyFolder &DrawRef=Main.DrawSettings;

};

// This measure is for controlling the rotation.
AnyKinRotational PedalJointRot = {
AnyRefFrame &Ground = …BikeModel.BikeFrame;
AnyRefNode &Foot = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;
Type=RotAxesAngles;
};

Hopefully we’ve done some foolish error that is easy to fix!

At current time our model is finding an optimum, with regards to
minimum metabolism, position for the cyclist at SaddleHeight = 0.56m
and SaddlePosition = -0.30m and a “comfortable” PedalArmWidth =
0.17m. The poor man looks like he’s half a second away from a large
accident.

Thanks on the clarification of the backstroke, too. We had hoped the
program would work in exactly that way.

Thank you once again for being brilliant,

Henrik
Tobias

— In anyscript@yahoogroups.com, “AnyBody Support” <support@…>
wrote:
>
> Hi Henrik and Tobias
>
>
>
> I am a little uncertain about the first question, but i will try to
explain
> how it works.
>
> The crank torque is defined in the file “BikeFrameAndWheel.any” by
the
> following lines:
>
>
>
> //This is the definiton of the crank moment
>
> AnyForce CrankMoment ={
>
> AnyJoint &Jnt=.CrankJoint;
>
>
> F={…BikeParameters.CrankMomentOffset-
…BikeParameters.CrankMomentAmpcos(
> 4pit/…BikeParameters.T+ …BikeParameters.CrankMomentPhase)};
>
> };
>
>
>
>
>
> As you can see the torque is related to the parameters in the main
file so
> these settings control the torque.
>
>
>
> In the same file you will also find the definition of the crank
motion
>
> AnyKinEqSimpleDriver CrankRotation = {
>
> AnyRevoluteJoint &Jnt = .CrankJoint;
>
> DriverPos = {0};
>
> DriverVel = {-(…BikeParameters.Cadence/60)pi2};
>
> Reaction.Type = {Off}; //This has to be zero otherwise the
crank driver
> will carry the applied load
>
> };
>
>
>
> Again you can see that it directly refers to the parameters of the
main
> file.
>
>
>
> When looking at the ankle motion this is as you know controlled by
a fourier
> driver in the JointsAndDrivers.any file please see
>
>
>
> AnyKinEqFourierDriver PedalRRotationFourier ={
>
> AnyKinMeasureOrg &Jnt = .FootRMovement;
>
> Type = CosSin;
>
> AnyVar a0 = -0.3797-0.5pi;
>
> AnyVar a1 = -0.0205;
>
> AnyVar b1 = 0.3393;
>
> AnyVar a2 = 0.004906;
>
> AnyVar b2 = 0.03654;
>
> AnyVar w = 7.681;
>
> Freq = …BikeParameters.Cadence/60;
>
> A = { {a0+0.5*pi,a1,a2} };
>
> B = {{0,b1,b2} };
>
> Reaction.Type = {Off};
>
> };
>
>
>
> This motion is only linked to the parameters of the main file in
terms of
> the cadency.
>
>
>
> Concerning the foot pedal joint I tried to dumb the sRel values of
the two
> nodes ToeJoint and GroundJoint you mention, this is the result:
>
> -------------------------------Objects dump-------------------------

>
> Main.Model.HumanModel.Right.Leg.Seg.Foot.ToeJoint.sRel = {-0, -
0.0671,
> 0.025};
>
> -------------------------------Objects dump-------------------------

>
> Main.Model.HumanModel.Right.Leg.Seg.Foot.GroundJoint.sRel = {0, 0,
0};
>
>
>
> There is a big difference in these nodes so shifting the node in
the joint
> should be visible once the kinematic has been solved (not a load
time!)
>
>
>
> I tried to do this change and it worked as intended, the joint
moved to the
> front of the foot. Below is the modified joints for the right side
>
>
>
> AnyRevoluteJoint PedalFootJoint = {
>
> AnyRefNode &ref1= …BikeModel.Crank.RPedal;
>
> AnyRefNode &ref2 = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;
>
> #include
> "…/…/…/BRep/Aalborg/ToolBox/JointReactions/DrawJointReactions.any
"
>
> AnyFolder &DrawRef=Main.DrawSettings;
>
> };
>
>
>
> // This measure is for controlling the rotation.
>
> AnyKinRotational PedalJointRot = {
>
> AnyRefFrame &Ground = …BikeModel.BikeFrame;
>
> AnyRefNode &Foot = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;
>
> Type=RotAxesAngles;
>
> };
>
>
>
>
>
> About the backstroke: If it will be of any advantage in terms of
lowering
> the activities of the muscles the model will pull the leg upward at
the
> backstroke. There is nothing in the model that prevents this from
happing;
> the foot is attached to the pedal through a revolute joint which
can also
> carry such a pulling load. So if the model does not show this it
is likely
> to be related to the loading conditions and motion.
>
>
>
> I hope this answered the questions otherwise please ask again.
>
>
>
> Best regards
>
> SÃ¸ren, anybody Support
>
>
>
>
>
>
>
> _____
>
> From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com]
On Behalf
> Of vtveteran
> Sent: 29 November 2007 15:59
> To: anyscript@yahoogroups.com
> Subject: [AnyScript] Re: Trouble with multi-dimensional analysis
and home
> installation
>
>
>
> Hi John!
>
> Now we feel stupid. We trawled all through the JointsAndDrivers
> file of the OptimStudy looking for differences to the one in
Bike3D,
> and only finding slashed-out text adding an extra dimension to the
A
> and B variables in the Fourier driver. It never occurred to us that
> it could be something so simple as a variable in the main file.
> Anyway, I’m curious as to how this is implemented. It feels odd
that
> we merely state these values (the phase shift and all the wiggles)
as
> variables in the main file and then they magically perform their
> function. Don’t we have to implement them into the method somehow?
>
> We also tried to change the foot position on the pedal by switching
> from GroundJoint to ToeJoint in the definition of both the foot-
joint
> connection and the line of rotation of the foot-pedal. This has
> however had no effect whatsoever on the image that appears in the
> model window; the forces on the pedals (we presume the green arrows
> that vary throughout the cycle are force avatars) still eminate
from
> the centre of the foot, where we don’t want them!!
>
> We would also like to know more about the mechanism rotating the
> pedal when the foot is not on the down stroke. The force arrows, if
> that is what they are, show no evidence of the leg pulling the
pedal
> up on the backstroke, as professional cyclists do. Is this feature
> implemented into the model? Given the foot placement on the pedal
and
> the low default mechanical output of 170W, we presume that the
model
> is more of bike-as-transport cyclist, rather than a professional
> racer, which is what we are looking to model. We presume that a pro
> is capable of putting out a much larger CrankMomentTopDeadCentre
than
> 1.5, but we can’t seem to find any literature on that anywhere.
>
> Thanks for all your help so far, it is very much appreciated!
>
> Greetings
>
> Henrik SjÃ¶stedt
> Tobias Lestrell
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com,
> “AnyBody Support” <support@>
> wrote:
> >
> > Wow, the group’s first posting in Swedish! Luckily we AnyBuddies
> are
> > very multilingual, but I’ll answer in English anyway for the
> benefit
> > of the other users.
> >
> > The construction of the Foruier driver function is pretty much
> > structured like you will find it in any text book on mathematical
> > analysis. However, rather than me explaining a whole lot, please
> have
> > a look at the “Parameter studies and optimization” tutorial here:
> > http://www.anybodyt <http://www.anybodytech.com/654.0.html.>
> ech.com/654.0.html.
> >
> > This tutorial uses a bicycle model where the joint torque is
phase
> > shifted just like you want. You can download the model from a
link
> > inside the tutorial.
> >
> > The phase shift is accomplished in the main file by the following
> > line:
> > AnyVar PhaseShift = atan(SaddlePos/SaddleHeight);
> >
> > The variable PhaseShift is subsequently used in the definition of
> the
> > crank torque.
> >
> > Best regards,
> > John
> >
> > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com,
> “vtveteran” <henriksjostedt@>
> > wrote:
> > >
> > > Hej John!
> > >
> > > Nu har vi fler problem:
> > >
> > > Vi har fÃ¶rstÃ¥tt att b1,b2 ger B i Fourierdrivern. De Ã¤r satta
> till
> > > konstanter. Ã„r det de som vi skall Ã¤ndra?
> > > Finns det nÃ¥gon ledtrÃ¥d i nÃ¥got som ni har skrivit om hur vi
> skall
> > > bestÃ¤mma dessa konstanter? Hur funkar det?
> > >
> > > Sen undrar vi ocksÃ¥ hur foten sitter fast i pedalen?
> > > Tror du det Ã¤r smart att lÃ¤gga in en extra nod pÃ¥ foten som vi
> > > fÃ¶rbinder med pedalen?
> > > Det ser lite konstigt ut pÃ¥ modellen bilden. Det ser ut som han
> > > trampar med hÃ¥lfoten!
> > >
> > > HÃ¤lsningar
> > > Ingrid
> > > med flera
> > >
> >
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

system · December 3, 2007, 9:29am

Hi Henrik and Tobias

Thanks for the kind words.

Your AnyScript for connecting the right foot toejoint with the pedal looks
ok, i can not spot any errors there…

I can think of two things which would cause for you problem

1 The file you are doing this change in, are for some reason not part of the
model anymore? Please check that it says “loaded” when you are loading the
model

2 You may not have been running any analysis yet… at loading time you will
not see the foot connected to this new point.

It sounds a bit strange with the seat position it may be related to the foot
position on the pedals and the applied crankmoment?

Best regards

SÃ¸ren

From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of vtveteran
Sent: 30 November 2007 15:37
To: anyscript@yahoogroups.com
Subject: [AnyScript] Re: Trouble with multi-dimensional analysis and home
installation

Hi SÃ¶ren!

You guys really are fantastic when it comes to helping us, thanks a
lot!

Sadly, we still can’t get the pedal connection to shift on the foot.
The lines in our Leg.Seg folder relating to the problem are:

AnyRefNode Toetip = {
sRel = .Scale({-0.012,-0.126,0.023});
};

AnyRefNode GroundJoint = {
sRel = .Scale({0.0, 0.0, …Sign*0.0});
ARel = {{0, 1, 0}, {-1, 0, 0}, {0, 0, 1}};
};

AnyRefNode HeelJoint = {
sRel = .Scale({0.009, 0.115, …Sign*0.00});
};

AnyRefNode ToeJoint = {
sRel = .Scale({-0.0, -0.0671, …Sign*0.025});
ARel = {{0, 1, 0}, {-1, 0, 0}, {0, 0, 1}};
};

and the lines for the right hand side in JointsAndDrivers:

AnyFolder Right={
// This is the joint connecting the foot to the pedal. It simply
// specifies that the linear distances between the two must be zero

/AnyKinLinear PedalJointLin = {
AnyRefNode &PedalR= …BikeModel.Crank.RPedal;
AnyRefNode &Foot = …ModelFolder.Right.Leg.Seg.Foot.GroundJoint;
};/

AnyRevoluteJoint PedalFootJoint = {
AnyRefNode &ref1= …BikeModel.Crank.RPedal;
AnyRefNode &ref2 = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;

#include “…/…/…/BRep/Aalborg/ToolBox/JointReactions/DrawJointReacti
ons.any”

AnyFolder &DrawRef=Main.DrawSettings;

};

// This measure is for controlling the rotation.
AnyKinRotational PedalJointRot = {
AnyRefFrame &Ground = …BikeModel.BikeFrame;
AnyRefNode &Foot = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;
Type=RotAxesAngles;
};

Hopefully we’ve done some foolish error that is easy to fix!

At current time our model is finding an optimum, with regards to
minimum metabolism, position for the cyclist at SaddleHeight = 0.56m
and SaddlePosition = -0.30m and a “comfortable” PedalArmWidth =
0.17m. The poor man looks like he’s half a second away from a large
accident.

Thanks on the clarification of the backstroke, too. We had hoped the
program would work in exactly that way.

Thank you once again for being brilliant,

Henrik
Tobias

— In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
“AnyBody Support” <support@…>
wrote:
>
> Hi Henrik and Tobias
>
>
>
> I am a little uncertain about the first question, but i will try to
explain
> how it works.
>
> The crank torque is defined in the file “BikeFrameAndWheel.any” by
the
> following lines:
>
>
>
> //This is the definiton of the crank moment
>
> AnyForce CrankMoment ={
>
> AnyJoint &Jnt=.CrankJoint;
>
>
> F={…BikeParameters.CrankMomentOffset-
…BikeParameters.CrankMomentAmpcos(
> 4pit/…BikeParameters.T+ …BikeParameters.CrankMomentPhase)};
>
> };
>
>
>
>
>
> As you can see the torque is related to the parameters in the main
file so
> these settings control the torque.
>
>
>
> In the same file you will also find the definition of the crank
motion
>
> AnyKinEqSimpleDriver CrankRotation = {
>
> AnyRevoluteJoint &Jnt = .CrankJoint;
>
> DriverPos = {0};
>
> DriverVel = {-(…BikeParameters.Cadence/60)pi2};
>
> Reaction.Type = {Off}; //This has to be zero otherwise the
crank driver
> will carry the applied load
>
> };
>
>
>
> Again you can see that it directly refers to the parameters of the
main
> file.
>
>
>
> When looking at the ankle motion this is as you know controlled by
a fourier
> driver in the JointsAndDrivers.any file please see
>
>
>
> AnyKinEqFourierDriver PedalRRotationFourier ={
>
> AnyKinMeasureOrg &Jnt = .FootRMovement;
>
> Type = CosSin;
>
> AnyVar a0 = -0.3797-0.5pi;
>
> AnyVar a1 = -0.0205;
>
> AnyVar b1 = 0.3393;
>
> AnyVar a2 = 0.004906;
>
> AnyVar b2 = 0.03654;
>
> AnyVar w = 7.681;
>
> Freq = …BikeParameters.Cadence/60;
>
> A = { {a0+0.5*pi,a1,a2} };
>
> B = {{0,b1,b2} };
>
> Reaction.Type = {Off};
>
> };
>
>
>
> This motion is only linked to the parameters of the main file in
terms of
> the cadency.
>
>
>
> Concerning the foot pedal joint I tried to dumb the sRel values of
the two
> nodes ToeJoint and GroundJoint you mention, this is the result:
>
> -------------------------------Objects dump-------------------------

>
> Main.Model.HumanModel.Right.Leg.Seg.Foot.ToeJoint.sRel = {-0, -
0.0671,
> 0.025};
>
> -------------------------------Objects dump-------------------------

>
> Main.Model.HumanModel.Right.Leg.Seg.Foot.GroundJoint.sRel = {0, 0,
0};
>
>
>
> There is a big difference in these nodes so shifting the node in
the joint
> should be visible once the kinematic has been solved (not a load
time!)
>
>
>
> I tried to do this change and it worked as intended, the joint
moved to the
> front of the foot. Below is the modified joints for the right side
>
>
>
> AnyRevoluteJoint PedalFootJoint = {
>
> AnyRefNode &ref1= …BikeModel.Crank.RPedal;
>
> AnyRefNode &ref2 = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;
>
> #include
> "…/…/…/BRep/Aalborg/ToolBox/JointReactions/DrawJointReactions.any
"
>
> AnyFolder &DrawRef=Main.DrawSettings;
>
> };
>
>
>
> // This measure is for controlling the rotation.
>
> AnyKinRotational PedalJointRot = {
>
> AnyRefFrame &Ground = …BikeModel.BikeFrame;
>
> AnyRefNode &Foot = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;
>
> Type=RotAxesAngles;
>
> };
>
>
>
>
>
> About the backstroke: If it will be of any advantage in terms of
lowering
> the activities of the muscles the model will pull the leg upward at
the
> backstroke. There is nothing in the model that prevents this from
happing;
> the foot is attached to the pedal through a revolute joint which
can also
> carry such a pulling load. So if the model does not show this it
is likely
> to be related to the loading conditions and motion.
>
>
>
> I hope this answered the questions otherwise please ask again.
>
>
>
> Best regards
>
> SÃ¸ren, anybody Support
>
>
>
>
>
>
>
> _____
>
> From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
[mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com]
On Behalf
> Of vtveteran
> Sent: 29 November 2007 15:59
> To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> Subject: [AnyScript] Re: Trouble with multi-dimensional analysis
and home
> installation
>
>
>
> Hi John!
>
> Now we feel stupid. We trawled all through the JointsAndDrivers
> file of the OptimStudy looking for differences to the one in
Bike3D,
> and only finding slashed-out text adding an extra dimension to the
A
> and B variables in the Fourier driver. It never occurred to us that
> it could be something so simple as a variable in the main file.
> Anyway, I’m curious as to how this is implemented. It feels odd
that
> we merely state these values (the phase shift and all the wiggles)
as
> variables in the main file and then they magically perform their
> function. Don’t we have to implement them into the method somehow?
>
> We also tried to change the foot position on the pedal by switching
> from GroundJoint to ToeJoint in the definition of both the foot-
joint
> connection and the line of rotation of the foot-pedal. This has
> however had no effect whatsoever on the image that appears in the
> model window; the forces on the pedals (we presume the green arrows
> that vary throughout the cycle are force avatars) still eminate
from
> the centre of the foot, where we don’t want them!!
>
> We would also like to know more about the mechanism rotating the
> pedal when the foot is not on the down stroke. The force arrows, if
> that is what they are, show no evidence of the leg pulling the
pedal
> up on the backstroke, as professional cyclists do. Is this feature
> implemented into the model? Given the foot placement on the pedal
and
> the low default mechanical output of 170W, we presume that the
model
> is more of bike-as-transport cyclist, rather than a professional
> racer, which is what we are looking to model. We presume that a pro
> is capable of putting out a much larger CrankMomentTopDeadCentre
than
> 1.5, but we can’t seem to find any literature on that anywhere.
>
> Thanks for all your help so far, it is very much appreciated!
>
> Greetings
>
> Henrik SjÃ¶stedt
> Tobias Lestrell
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com,
> “AnyBody Support” <support@>
> wrote:
> >
> > Wow, the group’s first posting in Swedish! Luckily we AnyBuddies
> are
> > very multilingual, but I’ll answer in English anyway for the
> benefit
> > of the other users.
> >
> > The construction of the Foruier driver function is pretty much
> > structured like you will find it in any text book on mathematical
> > analysis. However, rather than me explaining a whole lot, please
> have
> > a look at the “Parameter studies and optimization” tutorial here:
> > http://www.anybodyt <http://www.anybodyt
<http://www.anybodytech.com/654.0.html.> ech.com/654.0.html.>
> ech.com/654.0.html.
> >
> > This tutorial uses a bicycle model where the joint torque is
phase
> > shifted just like you want. You can download the model from a
link
> > inside the tutorial.
> >
> > The phase shift is accomplished in the main file by the following
> > line:
> > AnyVar PhaseShift = atan(SaddlePos/SaddleHeight);
> >
> > The variable PhaseShift is subsequently used in the definition of
> the
> > crank torque.
> >
> > Best regards,
> > John
> >
> > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com,
> “vtveteran” <henriksjostedt@>
> > wrote:
> > >
> > > Hej John!
> > >
> > > Nu har vi fler problem:
> > >
> > > Vi har fÃ¶rstÃ¥tt att b1,b2 ger B i Fourierdrivern. De Ã¤r satta
> till
> > > konstanter. Ã„r det de som vi skall Ã¤ndra?
> > > Finns det nÃ¥gon ledtrÃ¥d i nÃ¥got som ni har skrivit om hur vi
> skall
> > > bestÃ¤mma dessa konstanter? Hur funkar det?
> > >
> > > Sen undrar vi ocksÃ¥ hur foten sitter fast i pedalen?
> > > Tror du det Ã¤r smart att lÃ¤gga in en extra nod pÃ¥ foten som vi
> > > fÃ¶rbinder med pedalen?
> > > Det ser lite konstigt ut pÃ¥ modellen bilden. Det ser ut som han
> > > trampar med hÃ¥lfoten!
> > >
> > > HÃ¤lsningar
> > > Ingrid
> > > med flera
> > >
> >
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

[Non-text portions of this message have been removed]

system · December 6, 2007, 5:40pm

Hi guys,

We got the foot to shift. We were indeed modifying an old copy of the
JointsAndDrivers file that wasn’t being loaded. He now has a much more
satisfactory foot position. Unfortunately, it hasn’t led to as much
change in our results as we would have liked. We are getting an
optimal result for the SaddlePosition most times we run now, but the
SaddleHeight variable still tends to the lowest possible value, which
we struggle to believe can actually be correct. Unless, of course, the
fellow has worked out that recumbent bikes are more efficient and is
trying to ape one. At this stage, we really are rather out of ideas as
to what could be wrong. We applied the phase change and corresponding
wiggles to the main file like this:

Main = {

AnyFolder BikeParameters = {

 // Geometry parameters
 AnyVar PedalArmLength =0.17;  //Length of pedal arm
 AnyVar PedalArmWidth = 0.10; //Horizontal distance between left

and right connecting point between foot and pedal
AnyVar SaddleHeight = 0.72; //Height of hip joint measured
vertically from the crank
AnyVar SaddlePos = -0.15; //Horizontal pos of hipjoint
measured from the crank

 // Loading parameters
 AnyVar Cadence = 80.0;   //Cadence in RPM
 AnyVar MechOutput = 500; //Average Mechanical output over a cycle

in Watts
AnyVar PhaseShift = atan(SaddlePos/SaddleHeight); // Phase shift
for different seat positions
AnyVar WigglePhase = 0.60-PhaseShift; // Phase shift for different
seat positions
AnyVar FootWiggleAmp = 20pi/180; // Amplitude of ankle
dorsi/plantar flextion w.r.t. ground
AnyVar FootWiggleOffs = -23pi/180; // Offset of dorsi/plantar
flextion

 // The function for the crank moment is defined as

Moment=Offset-Ampcos(4pi*t/T+Phase)
AnyVar T = 60/Cadence; //Cycle time

 AnyVar CrankMomentTopDeadCenter = 3; // This is the moment which

can be applied by the rider in the top dead center point
AnyVar CrankMomentOffset = (MechOutputT)/(2pi);
AnyVar CrankMomentAmp = CrankMomentOffset-CrankMomentTopDeadCenter;
AnyVar CrankMomentPhase = -PhaseShift;

 AnyVar DistBetweenBikes =1;  // Physical placement of the two models

};

I hope we did it wrong, because then at least there would be something
to fix. We would love it if you guys had any other ideas or thoughts
about the whole thing. AnyBody experts are rather thin on the ground
in Lund so there aren’t very many other people we can ask for help.

Thanks once again for all your kind help.

Henrik SjÃ¶stedt
Tobias Lestrell

— In anyscript@yahoogroups.com, “AnyBody Support” <support@…> wrote:
>
> Hi Henrik and Tobias
>
>
>
> Thanks for the kind words.
>
>
>
> Your AnyScript for connecting the right foot toejoint with the pedal
looks
> ok, i can not spot any errors there…
>
>
>
> I can think of two things which would cause for you problem
>
> 1 The file you are doing this change in, are for some reason not
part of the
> model anymore? Please check that it says “loaded” when you are
loading the
> model
>
> 2 You may not have been running any analysis yet… at loading time
you will
> not see the foot connected to this new point.
>
>
>
> It sounds a bit strange with the seat position it may be related to
the foot
> position on the pedals and the applied crankmoment?
>
>
>
> Best regards
>
>
>
> SÃ¸ren
>
>
>
> _____
>
> From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com]
On Behalf
> Of vtveteran
> Sent: 30 November 2007 15:37
> To: anyscript@yahoogroups.com
> Subject: [AnyScript] Re: Trouble with multi-dimensional analysis and
home
> installation
>
>
>
> Hi SÃ¶ren!
>
> You guys really are fantastic when it comes to helping us, thanks a
> lot!
>
> Sadly, we still can’t get the pedal connection to shift on the foot.
> The lines in our Leg.Seg folder relating to the problem are:
>
> AnyRefNode Toetip = {
> sRel = .Scale({-0.012,-0.126,0.023});
> };
>
> AnyRefNode GroundJoint = {
> sRel = .Scale({0.0, 0.0, …Sign0.0});
> ARel = {{0, 1, 0}, {-1, 0, 0}, {0, 0, 1}};
> };
>
> AnyRefNode HeelJoint = {
> sRel = .Scale({0.009, 0.115, …Sign0.00});
> };
>
> AnyRefNode ToeJoint = {
> sRel = .Scale({-0.0, -0.0671, …Sign0.025});
> ARel = {{0, 1, 0}, {-1, 0, 0}, {0, 0, 1}};
> };
>
> and the lines for the right hand side in JointsAndDrivers:
>
> AnyFolder Right={
> // This is the joint connecting the foot to the pedal. It simply
> // specifies that the linear distances between the two must be zero
>
> /AnyKinLinear PedalJointLin = {
> AnyRefNode &PedalR= …BikeModel.Crank.RPedal;
> AnyRefNode &Foot = …ModelFolder.Right.Leg.Seg.Foot.GroundJoint;
> };/
>
> AnyRevoluteJoint PedalFootJoint = {
> AnyRefNode &ref1= …BikeModel.Crank.RPedal;
> AnyRefNode &ref2 = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;
>
> #include “…/…/…/BRep/Aalborg/ToolBox/JointReactions/DrawJointReacti
> ons.any”
>
> AnyFolder &DrawRef=Main.DrawSettings;
>
> };
>
> // This measure is for controlling the rotation.
> AnyKinRotational PedalJointRot = {
> AnyRefFrame &Ground = …BikeModel.BikeFrame;
> AnyRefNode &Foot = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;
> Type=RotAxesAngles;
> };
>
> Hopefully we’ve done some foolish error that is easy to fix!
>
> At current time our model is finding an optimum, with regards to
> minimum metabolism, position for the cyclist at SaddleHeight = 0.56m
> and SaddlePosition = -0.30m and a “comfortable” PedalArmWidth =
> 0.17m. The poor man looks like he’s half a second away from a large
> accident.
>
> Thanks on the clarification of the backstroke, too. We had hoped the
> program would work in exactly that way.
>
> Thank you once again for being brilliant,
>
> Henrik
> Tobias
>
> — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com,
> “AnyBody Support” <support@>
> wrote:
> >
> > Hi Henrik and Tobias
> >
> >
> >
> > I am a little uncertain about the first question, but i will try to
> explain
> > how it works.
> >
> > The crank torque is defined in the file “BikeFrameAndWheel.any” by
> the
> > following lines:
> >
> >
> >
> > //This is the definiton of the crank moment
> >
> > AnyForce CrankMoment ={
> >
> > AnyJoint &Jnt=.CrankJoint;
> >
> >
> > F={…BikeParameters.CrankMomentOffset-
> …BikeParameters.CrankMomentAmpcos(
> > 4pit/…BikeParameters.T+ …BikeParameters.CrankMomentPhase)};
> >
> > };
> >
> >
> >
> >
> >
> > As you can see the torque is related to the parameters in the main
> file so
> > these settings control the torque.
> >
> >
> >
> > In the same file you will also find the definition of the crank
> motion
> >
> > AnyKinEqSimpleDriver CrankRotation = {
> >
> > AnyRevoluteJoint &Jnt = .CrankJoint;
> >
> > DriverPos = {0};
> >
> > DriverVel = {-(…BikeParameters.Cadence/60)pi2};
> >
> > Reaction.Type = {Off}; //This has to be zero otherwise the
> crank driver
> > will carry the applied load
> >
> > };
> >
> >
> >
> > Again you can see that it directly refers to the parameters of the
> main
> > file.
> >
> >
> >
> > When looking at the ankle motion this is as you know controlled by
> a fourier
> > driver in the JointsAndDrivers.any file please see
> >
> >
> >
> > AnyKinEqFourierDriver PedalRRotationFourier ={
> >
> > AnyKinMeasureOrg &Jnt = .FootRMovement;
> >
> > Type = CosSin;
> >
> > AnyVar a0 = -0.3797-0.5pi;
> >
> > AnyVar a1 = -0.0205;
> >
> > AnyVar b1 = 0.3393;
> >
> > AnyVar a2 = 0.004906;
> >
> > AnyVar b2 = 0.03654;
> >
> > AnyVar w = 7.681;
> >
> > Freq = …BikeParameters.Cadence/60;
> >
> > A = { {a0+0.5pi,a1,a2} };
> >
> > B = {{0,b1,b2} };
> >
> > Reaction.Type = {Off};
> >
> > };
> >
> >
> >
> > This motion is only linked to the parameters of the main file in
> terms of
> > the cadency.
> >
> >
> >
> > Concerning the foot pedal joint I tried to dumb the sRel values of
> the two
> > nodes ToeJoint and GroundJoint you mention, this is the result:
> >
> > -------------------------------Objects dump-------------------------
> -
> >
> > Main.Model.HumanModel.Right.Leg.Seg.Foot.ToeJoint.sRel = {-0, -
> 0.0671,
> > 0.025};
> >
> > -------------------------------Objects dump-------------------------
> -
> >
> > Main.Model.HumanModel.Right.Leg.Seg.Foot.GroundJoint.sRel = {0, 0,
> 0};
> >
> >
> >
> > There is a big difference in these nodes so shifting the node in
> the joint
> > should be visible once the kinematic has been solved (not a load
> time!)
> >
> >
> >
> > I tried to do this change and it worked as intended, the joint
> moved to the
> > front of the foot. Below is the modified joints for the right side
> >
> >
> >
> > AnyRevoluteJoint PedalFootJoint = {
> >
> > AnyRefNode &ref1= …BikeModel.Crank.RPedal;
> >
> > AnyRefNode &ref2 = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;
> >
> > #include
> > "…/…/…/BRep/Aalborg/ToolBox/JointReactions/DrawJointReactions.any
> "
> >
> > AnyFolder &DrawRef=Main.DrawSettings;
> >
> > };
> >
> >
> >
> > // This measure is for controlling the rotation.
> >
> > AnyKinRotational PedalJointRot = {
> >
> > AnyRefFrame &Ground = …BikeModel.BikeFrame;
> >
> > AnyRefNode &Foot = …ModelFolder.Right.Leg.Seg.Foot.ToeJoint;
> >
> > Type=RotAxesAngles;
> >
> > };
> >
> >
> >
> >
> >
> > About the backstroke: If it will be of any advantage in terms of
> lowering
> > the activities of the muscles the model will pull the leg upward at
> the
> > backstroke. There is nothing in the model that prevents this from
> happing;
> > the foot is attached to the pedal through a revolute joint which
> can also
> > carry such a pulling load. So if the model does not show this it
> is likely
> > to be related to the loading conditions and motion.
> >
> >
> >
> > I hope this answered the questions otherwise please ask again.
> >
> >
> >
> > Best regards
> >
> > SÃ¸ren, anybody Support
> >
> >
> >
> >
> >
> >
> >
> > _____
> >
> > From: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> [mailto:anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
ps.com]
> On Behalf
> > Of vtveteran
> > Sent: 29 November 2007 15:59
> > To: anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com> ps.com
> > Subject: [AnyScript] Re: Trouble with multi-dimensional analysis
> and home
> > installation
> >
> >
> >
> > Hi John!
> >
> > Now we feel stupid. We trawled all through the JointsAndDrivers
> > file of the OptimStudy looking for differences to the one in
> Bike3D,
> > and only finding slashed-out text adding an extra dimension to the
> A
> > and B variables in the Fourier driver. It never occurred to us that
> > it could be something so simple as a variable in the main file.
> > Anyway, I’m curious as to how this is implemented. It feels odd
> that
> > we merely state these values (the phase shift and all the wiggles)
> as
> > variables in the main file and then they magically perform their
> > function. Don’t we have to implement them into the method somehow?
> >
> > We also tried to change the foot position on the pedal by switching
> > from GroundJoint to ToeJoint in the definition of both the foot-
> joint
> > connection and the line of rotation of the foot-pedal. This has
> > however had no effect whatsoever on the image that appears in the
> > model window; the forces on the pedals (we presume the green arrows
> > that vary throughout the cycle are force avatars) still eminate
> from
> > the centre of the foot, where we don’t want them!!
> >
> > We would also like to know more about the mechanism rotating the
> > pedal when the foot is not on the down stroke. The force arrows, if
> > that is what they are, show no evidence of the leg pulling the
> pedal
> > up on the backstroke, as professional cyclists do. Is this feature
> > implemented into the model? Given the foot placement on the pedal
> and
> > the low default mechanical output of 170W, we presume that the
> model
> > is more of bike-as-transport cyclist, rather than a professional
> > racer, which is what we are looking to model. We presume that a pro
> > is capable of putting out a much larger CrankMomentTopDeadCentre
> than
> > 1.5, but we can’t seem to find any literature on that anywhere.
> >
> > Thanks for all your help so far, it is very much appreciated!
> >
> > Greetings
> >
> > Henrik SjÃ¶stedt
> > Tobias Lestrell
> >
> > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com,
> > “AnyBody Support” <support@>
> > wrote:
> > >
> > > Wow, the group’s first posting in Swedish! Luckily we AnyBuddies
> > are
> > > very multilingual, but I’ll answer in English anyway for the
> > benefit
> > > of the other users.
> > >
> > > The construction of the Foruier driver function is pretty much
> > > structured like you will find it in any text book on mathematical
> > > analysis. However, rather than me explaining a whole lot, please
> > have
> > > a look at the “Parameter studies and optimization” tutorial here:
> > > http://www.anybodyt <http://www.anybodyt
> <http://www.anybodytech.com/654.0.html.> ech.com/654.0.html.>
> > ech.com/654.0.html.
> > >
> > > This tutorial uses a bicycle model where the joint torque is
> phase
> > > shifted just like you want. You can download the model from a
> link
> > > inside the tutorial.
> > >
> > > The phase shift is accomplished in the main file by the following
> > > line:
> > > AnyVar PhaseShift = atan(SaddlePos/SaddleHeight);
> > >
> > > The variable PhaseShift is subsequently used in the definition of
> > the
> > > crank torque.
> > >
> > > Best regards,
> > > John
> > >
> > > — In anyscript@yahoogrou <mailto:anyscript%40yahoogroups.com>
> ps.com,
> > “vtveteran” <henriksjostedt@>
> > > wrote:
> > > >
> > > > Hej John!
> > > >
> > > > Nu har vi fler problem:
> > > >
> > > > Vi har fÃ¶rstÃ¥tt att b1,b2 ger B i Fourierdrivern. De Ã¤r satta
> > till
> > > > konstanter. Ã„r det de som vi skall Ã¤ndra?
> > > > Finns det nÃ¥gon ledtrÃ¥d i nÃ¥got som ni har skrivit om hur vi
> > skall
> > > > bestÃ¤mma dessa konstanter? Hur funkar det?
> > > >
> > > > Sen undrar vi ocksÃ¥ hur foten sitter fast i pedalen?
> > > > Tror du det Ã¤r smart att lÃ¤gga in en extra nod pÃ¥ foten som vi
> > > > fÃ¶rbinder med pedalen?
> > > > Det ser lite konstigt ut pÃ¥ modellen bilden. Det ser ut som han
> > > > trampar med hÃ¥lfoten!
> > > >
> > > > HÃ¤lsningar
> > > > Ingrid
> > > > med flera
> > > >
> > >
> >
> >
> >
> >
> >
> > [Non-text portions of this message have been removed]
> >
>
>
>
>
>
> [Non-text portions of this message have been removed]
>

Trouble with multi-dimensional analysis and home installation

> > Main.Model.HumanModel.Right.Leg.Seg.Foot.ToeJoint.sRel = {-0, - 0.0671, > 0.025}; > > -------------------------------Objects dump-------------------------

> > Main.Model.HumanModel.Right.Leg.Seg.Foot.ToeJoint.sRel = {-0, - 0.0671, > 0.025}; > > -------------------------------Objects dump-------------------------

>
> Main.Model.HumanModel.Right.Leg.Seg.Foot.ToeJoint.sRel = {-0, -
0.0671,
> 0.025};
>
> -------------------------------Objects dump-------------------------

>
> Main.Model.HumanModel.Right.Leg.Seg.Foot.ToeJoint.sRel = {-0, -
0.0671,
> 0.025};
>
> -------------------------------Objects dump-------------------------