hi,
i want to model a human with backpack during walking.
i have seen the clip of your model in anybodetech.com and shoulder bag on repository 1.0.
i want to use shoulder bag file for modeling backpack and when i rotate the bag.stl file and load it, the new model is correct but when i run it, the bag rotates 90 degrees and go back to its initial position.
what should i do to prevent rotation?
i have another question:
which constraints should be eliminated to run correctly?
thanks
1 Like
Hi
i have a problem with backpack.
i added 2 perismatic joints to join the bag to body.
the program loads truely but when i want to run the program i encounter with the error “[SIZE=1]
ERROR(OBJ.MCH.KIN2)[/SIZE] : [SIZE=1]F:/AM…0/A…n/E…s/S…g/ShoulderBag.Main.any[/SIZE] : [SIZE=1]Study[/SIZE] : Model is kinematically over-constrained : Position analysis failed : 3 unsolvable constraint(s) found”
[SIZE=2]what should i do to solve this problem?
[/SIZE]
[SIZE=2][/SIZE]
[SIZE=2]thanks
[/SIZE]
Hi,
In general the kinematic constraints have to be kept in balance. So if you add some constraints somehwere to a working model (without adding any segment) you have to remove as much constraints somewhere else in the model to keep the balance. Otherwise the model is overconstrained will not run.
And for each segment added the six dof must be controlled by joint constraints or driver.
In the case of a backpack i would suggest to replace the hip prismatic joint by a new one on the back of the thorax, with a strap on each shoulder (but only one strap length should be driven for kinematic balance, the other will automatically follow).
As for the rotation i think there is no nned to rotate the stl file itself, as the stl file is attached to the bag segment, moving/rotating the segment will also move the stl file together.
Best regards, Sylvain.
thank you very much for your help.
do you mean that if i use an AnyKinSimpleDriver for strap1,then strap2 will automatically follow it?I mean that is there any other command that strap2 follows strap1?because i have the driver for strap1 and delete the strap2 driver and when i run the program the muscles of the side where i put strap 2 are activated and the muscles of the other side are passive. is that right?
thanks
Hi,
If you use a prismatic joint to link the bag to the thorax as suggested then in term of kinematic there is only one dof left to drive: the height of the bag. This equivalent to driving the length of one strap. If you drive the length of the two straps it means you drive two times the height of the bag and this would be over constrained.
So driving one strap is enough. There is no need for any kinematic constraint for the second strap.
However in term of forces it is diferent. I guess you want the load shared between the two straps. That means you need a reaction force on each strap. the first strap has a reaction force included with the driver (unless you specify it differently by using reaction.type=Off, which you should not). As the second strap has no driver, you have to applied an AnyReacForce to the Pline representing the strap.
That way both sides of the body will be loaded.
Best regards, sylvain.
very thanks for your help.
we modeled the gaitvaughan with hands,their muscles and drivers, and also trunk muscles. then added our backpack to its environment. but when we ran it, we encountered 3 warnings about muscle overloading. the types of our muscle are simple and we substituded them by 3E but our warning were more.
nStep in our model is 100. we lowered it to 22 and then the warnings were disappeared.
as an another solution, we changed SpecificMuscleTensionSpine and SpecificMuscleTensionShoulderArm in StrengthParameters folder from 150 to 300 and then we have no warning.
which solution is right? what is your suggestion about our problem?
Hi,
From your description the overloading seems to come from local picks in the muscle activity instead of a real general increase of the activity. If this is the case then it may be due to numeric. You can try changing the recruitment solver to MinMaxStrict and still keep the 100 steps, it may cures it.
You should be carrefull when changing the SpecificMuscleTensionSpine and SpecificMuscleTensionShoulderArm, this is basically a coefficient for the strength of all muscle of the spine or arms. So when you double it as you did it means your model is two times stronger than the average. It can be usefull to model sport athletes (generaly stronger than the average) or weak elderly persons, but care should be taken not to be excesive and model some kind of Superman.
In your case i would rather try to change the solver first before those parameters.
You should also be aware of some limitation of the model. In the GaitVaughan (as well as others gait model) the motion and ground reaction forces are recorded from a normal gait. A gait with back pack may be a little different in term of motion, and the ground reaction forces will for sure be different due to the weight of the bag. So those small inacurracies could also contribute in creating muscle overloading.
The StandingModel for example is not subject to such limitations because it calculates automaticaly the posture and ground reaction forces.
Best regards, Sylvain.
Hi,
Is there any way that you can give me the backpack file or some sort of demo like in the Balance Model video on Anybody’s youtube (http://www.youtube.com/watch?v=W7m4VzwEkU0)? I require it for a project and am having a hard adjusting things to form a proper model.
Thank you very much.
Hi,
The model from the animation on YouTube is actually something very old (from 2004) and it will take us a while to get it running in the current repository.
But we can of course help you setup your own model, what is the purpose? What is the target, you could use the StandingModel, add a new segment either as a fixed segment, or with straps on the shoulders and conditional contact on the back.
If you post your model, what you have so far, we can help you to find the next steps…