Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Bike cab simulation - Relative error greater than relative tolerance

Please login with a confirmed email address before reporting spam

Hi,

I´m trying to model a bike cab for my graduation project from university. I stated with a simple model based on the truss bridge tutorial so I can easily modify the tube sections and get fast solutions to test them.

However, I get the

"Failed to find a solution.
The relative error (2.4e+002) is greater than the relative tolerance.
Returned solution is not converged."

error.

I have tried modifying the mesh and checking boundary conditions but can't get it to find a solution.

I did model the cab later as a solid and got a nice solution with the same boundary conditions, so I'm not sure where the error comes from, maybe it's something simple that I'm over looking.

I tried to attach my file but the attach button doesn't seem to be working, let me know and maybe I can email it if someone can hae a look at it.

Thanks for any help!

10 Replies Last Post 2016年8月17日 GMT-4 22:15
Jeff Hiller COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 2016年8月3日 GMT-4 16:47
You want to use beams, not trusses. Trusses cannot take loading in their perpendicular direction via bending: all they can do is stretch or shorten: not a good representation of a metal tube used to make a bike frame.
Best,
Jeff
You want to use beams, not trusses. Trusses cannot take loading in their perpendicular direction via bending: all they can do is stretch or shorten: not a good representation of a metal tube used to make a bike frame. Best, Jeff

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 2016年8月3日 GMT-4 17:13
Thanks Jeff,

Although the tutorial is called Truss Bridge, it actually uses the beam interface along with shells (which I will also use farther down the road in a more complex model with some metal plates that are used to reinforce the structure).

Not sure if that's what you meant.

Thanks.
Thanks Jeff, Although the tutorial is called Truss Bridge, it actually uses the beam interface along with shells (which I will also use farther down the road in a more complex model with some metal plates that are used to reinforce the structure). Not sure if that's what you meant. Thanks.

Jeff Hiller COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 2016年8月3日 GMT-4 17:24
OK, my bad. My gut feeling remains that your problem is singular somehow (and therefore does not have a unique solution, leading to the solver's inability to converge to that unique solution). Have you checked that your boundary conditions remove all 6 rigid body motions?
Try attaching your model again after clearing the solution to reduce file size if needed.
Jeff
OK, my bad. My gut feeling remains that your problem is singular somehow (and therefore does not have a unique solution, leading to the solver's inability to converge to that unique solution). Have you checked that your boundary conditions remove all 6 rigid body motions? Try attaching your model again after clearing the solution to reduce file size if needed. Jeff

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 2016年8月3日 GMT-4 23:17
Hi,

No problem.

Tried uploading my file in different computers and at home but the "Attach File" button doesn't seem to work, it doesn't open any new window or any place to select the file, not sure why.

Anyway, I'm relatively new to COMSOL so, about the boundary conditions:
1) I added fixed constraints on 5 different points along the rear axle of the cab, I believe this should remove motion on all axes and also block rotation around the 'x' and 'y' axes (the rear axle runs along the 'z' axis)
2) Added prescribed displacement at the end of the front tube (known as headtube), it is set to zero on 'y' and 'z', allowing the endpoint to move only along 'x', besides being part of the load case, this should block rotation around 'z' axis.
3) Added loads at the top of the headtube, pedal axel and driver and passenger seats.

Let me know if you see anything is missing please, I will continue to investigate why I can't attach a file, maybe it's because this is my first post? Don't know.

Thanks again Jeff and anyone else who can help!
Hi, No problem. Tried uploading my file in different computers and at home but the "Attach File" button doesn't seem to work, it doesn't open any new window or any place to select the file, not sure why. Anyway, I'm relatively new to COMSOL so, about the boundary conditions: 1) I added fixed constraints on 5 different points along the rear axle of the cab, I believe this should remove motion on all axes and also block rotation around the 'x' and 'y' axes (the rear axle runs along the 'z' axis) 2) Added prescribed displacement at the end of the front tube (known as headtube), it is set to zero on 'y' and 'z', allowing the endpoint to move only along 'x', besides being part of the load case, this should block rotation around 'z' axis. 3) Added loads at the top of the headtube, pedal axel and driver and passenger seats. Let me know if you see anything is missing please, I will continue to investigate why I can't attach a file, maybe it's because this is my first post? Don't know. Thanks again Jeff and anyone else who can help!

Jeff Hiller COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 2016年8月4日 GMT-4 14:04
Hi Alfredo,
The "Attach" button seems to be misbehaving today on Firefox, but works on IE and Chrome. If you tried on Firefox earlier, can you try using one of those browsers to upload your file?
Sorry about this, I have reported the problem to our web team.
Best,
Jeff
Hi Alfredo, The "Attach" button seems to be misbehaving today on Firefox, but works on IE and Chrome. If you tried on Firefox earlier, can you try using one of those browsers to upload your file? Sorry about this, I have reported the problem to our web team. Best, Jeff

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 2016年8月4日 GMT-4 22:09
Hi,

I was in fact using Firefox and it works in Chrome.

So, I attach the file, you'll see it's a pretty simple case I believe.

Again, thanks and any help is appreciated,

Alfredo
Hi, I was in fact using Firefox and it works in Chrome. So, I attach the file, you'll see it's a pretty simple case I believe. Again, thanks and any help is appreciated, Alfredo


Jeff Hiller COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 2016年8月5日 GMT-4 17:32
It looks like something is wrong with your imported CAD file around point 6 and edge 7, and also the geometry looks like it may be missing an entire beam (Edge 42 has no equivalent on the other side of the frame: I am guessing that is not intentional).
Note that if you fix point 6 (where the problem seems to be, see above) the solution converges nicely.
Jeff.
It looks like something is wrong with your imported CAD file around point 6 and edge 7, and also the geometry looks like it may be missing an entire beam (Edge 42 has no equivalent on the other side of the frame: I am guessing that is not intentional). Note that if you fix point 6 (where the problem seems to be, see above) the solution converges nicely. Jeff.

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 2016年8月5日 GMT-4 17:59
Perfect, yes, I had overlooked that tiny edge (edge 7) that should not be there, I deleted point 6 and edge 7 and in fact the solution converged. The other missing edge is on purpose due to an assymetrical force that must be added later.

I will continue to add complexity to the model and come back if I find any obstacles hehe

Thank you very much for your help Jeff and have a great day,

Alfredo
Perfect, yes, I had overlooked that tiny edge (edge 7) that should not be there, I deleted point 6 and edge 7 and in fact the solution converged. The other missing edge is on purpose due to an assymetrical force that must be added later. I will continue to add complexity to the model and come back if I find any obstacles hehe Thank you very much for your help Jeff and have a great day, Alfredo

Jeff Hiller COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 2016年8月8日 GMT-4 09:06
The Attach button seems to work again now that Firefox has updated to version 48.0.
Jeff
The Attach button seems to work again now that Firefox has updated to version 48.0. Jeff

Please login with a confirmed email address before reporting spam

Posted: 8 years ago 2016年8月17日 GMT-4 22:15
Thanks again,

Could you take a look at this model where I try to incorporate the shell interfece? I'm not sure if I'm missing some border condition to make it converge or if I'm using the beam connections as I should.

Also, suppose that some of the edges that represents the shells (edge 28, the long one along the z axis and edges 98, 104, 113, 114 and 119 and similar) is not supposed to be a beam, so I'm not sure what data to enter there, it should not be in the beam interfece but I know I should imput some properties on it, right? I can't delete them either as they define the region that represents the shells.

But even if I imput them as beams (the one that says "with all edges as beams"), the model will still not converge, so not sure what's happening.

Besides the Truss Bridge tutorial, are there any other examples that ilustrate the combination of beams and shelss?

Tailwinds,

Alfredo
Thanks again, Could you take a look at this model where I try to incorporate the shell interfece? I'm not sure if I'm missing some border condition to make it converge or if I'm using the beam connections as I should. Also, suppose that some of the edges that represents the shells (edge 28, the long one along the z axis and edges 98, 104, 113, 114 and 119 and similar) is not supposed to be a beam, so I'm not sure what data to enter there, it should not be in the beam interfece but I know I should imput some properties on it, right? I can't delete them either as they define the region that represents the shells. But even if I imput them as beams (the one that says "with all edges as beams"), the model will still not converge, so not sure what's happening. Besides the Truss Bridge tutorial, are there any other examples that ilustrate the combination of beams and shelss? Tailwinds, Alfredo

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.