Inverse Dynamic Analysis error

Hello
my name is Kyle Imatani and I am a beginning user, have gone through
the tutorials provided, and am now working on my model for a static
full-body simulation (in a seated position also constrained at the
feet, and hands) . However, I have encountered the following problem
while attempting inverse dynamic analysis which I hope you can resolve.

ERROR(OBJ1) ‘Study’ : Muscle recruitment analysis failed, simplex
solver reports that solution does not satisfy all constraints

While I have balanced the appropriate number of constraints and dofs,
can set initial conditions and run kinematic analysis, inverse dynamic
analysis still fails. After perusing forums, I have attempted to add
.001 gravity in x and z directions, and have played with the
reaction.type on/off options. Unfortunately, I do not understand why
this error occurs or how to fix it. (I have attatched the model under
Imatani_model.zip)

I would really appreciate any help or advice on any of these questions
or topics. Thanks in advance,

Kyle Imatani

Hello Kyle,

Welcome to the group!

Change the Recruitment Solver to the MinMaxSimplex. This may help. Sometimes
one of the solvers outperforms the other.

Best regards,

Sebastian


From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of kyleimatani
Sent: 28. juli 2008 10:31
To: anyscript@yahoogroups.com
Subject: [AnyScript] Inverse Dynamic Analysis error

Hello
my name is Kyle Imatani and I am a beginning user, have gone through
the tutorials provided, and am now working on my model for a static
full-body simulation (in a seated position also constrained at the
feet, and hands) . However, I have encountered the following problem
while attempting inverse dynamic analysis which I hope you can resolve.

ERROR(OBJ1) ‘Study’ : Muscle recruitment analysis failed, simplex
solver reports that solution does not satisfy all constraints

While I have balanced the appropriate number of constraints and dofs,
can set initial conditions and run kinematic analysis, inverse dynamic
analysis still fails. After perusing forums, I have attempted to add
.001 gravity in x and z directions, and have played with the
reaction.type on/off options. Unfortunately, I do not understand why
this error occurs or how to fix it. (I have attatched the model under
Imatani_model.zip)

I would really appreciate any help or advice on any of these questions
or topics. Thanks in advance,

Kyle Imatani

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

Sebastin,
Thanks for the timely response! It turns out after further
inspection that I did not correctly constrain my environment objects
correctly (by switching reaction.types to on) in my environment.any
file. However, I will keep your response in mind as I move foward.

That being said I have a final question, for now :), that I hope you
could answer.

When/why exactly do you need to specify reaction.type to on? My
understanding is that when the body is supported in some manner in a
certain kinetic measure, you tell turn on the reaction.type forces.
So why does the environment object need this? Isnt giving it a kineq
constriant, or kineqsimpledriver enough? If you have a foot being
supported by the ground, but not in any other way, would the
reaction.type read (Off,On,Off) xyz?

Thanks again for the help
~kyle imatani

Hello Kyle,

You need to specify the Reaction.Type to “On” if you want to supply a
reaction force which will basically carry all the load. For example in the
joints this means that a kind of motor in brought in the system which will
generate the force needed, without needing any muscles. The AnyKineq class
deals with the kinematics of the system, therefore defining / constraining
the motion, but stating no information about the kinetics.

You have to make sure that all your segments are somehow supported. Either
direct or indirect through other connections to segments which are
“grounded”.

For example if the only contact of your model with the environment happens
to be in the foot all reaction should be turned on. Please see also the
tutorial

Best regards,

Sebastian


From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of kyleimatani
Sent: 29. juli 2008 04:13
To: anyscript@yahoogroups.com
Subject: [AnyScript] Re: Inverse Dynamic Analysis error

Sebastin,
Thanks for the timely response! It turns out after further
inspection that I did not correctly constrain my environment objects
correctly (by switching reaction.types to on) in my environment.any
file. However, I will keep your response in mind as I move foward.

That being said I have a final question, for now :), that I hope you
could answer.

When/why exactly do you need to specify reaction.type to on? My
understanding is that when the body is supported in some manner in a
certain kinetic measure, you tell turn on the reaction.type forces.
So why does the environment object need this? Isnt giving it a kineq
constriant, or kineqsimpledriver enough? If you have a foot being
supported by the ground, but not in any other way, would the
reaction.type read (Off,On,Off) xyz?

Thanks again for the help
~kyle imatani

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