• posted

Hi all,

I am working on a design project as a student in a University. We're designing the frame for a small formula FSAE car. I modeled the frame as a simple set of datum curves and points and analyzed them as beams. I was successful checking the stiffness in pro-M by applying a moment to the front and constraining the rear. My next step is to check the frame with the engine installed in place of some of the members. The motor is modeled as a part, so I opened a new assembly with the frame and motor.

My questions are these: What is the best way to model the connections between the motor, which is a 3-D feature, and the frame, which is a set of datum curves? I ran an analyses and it crashed. I assume this is because the motor and frame are only connected by datum planes in the assembly. I am hoping pro-M can analyze connections between the two without remodeling the frame as a 3-D model, or trying to replace the motor with a series of curves.

Thanks for any suggestions, Mike

• posted

: My questions are these: What is the best way to model the connections : between the motor, which is a 3-D feature, and the frame, which is a : set of datum curves? I ran an analyses and it crashed. I assume this : is because the motor and frame are only connected by datum planes in : the assembly.

This is probably a good guess. But, it could also be because you mixed curve and solid elements in a beam analysis. don't know how well ProM can handle this. You know, when you do beam analysis, it tries to collapse regular solid and surface features to mid-planes, loosing contact between assembled components which it also doesn't handle very well. In any case, get rid of planar constraints. In part mode, place points on the supporting curve structure in approximately the places where the motor will be connected. Place corresponding points on the motor block and use these for assembly. In fact, for beam analysis, instead of solid model, you should just use a flat surface with some points on it. Or, you might even try modelling the block as a wireframe outline with datum curves and place the points on those curves, as well. I suspect that the whole analysis will run better with similar construction of each part. And it will definitely run the fastest (and probably the most trouble-free) it can with all curves.

David Janes

• posted

Dear Mike,

If there is no geometry connection, you could try using 'rigid' connections. Only put the engine block in your frame model, as not much else contributes to stiffness. Place datum point on the curves where the engine mounts are (I assume that there are no rubber mountings otherwise you would use springs) and use a rigid to connect this point to the mounting surface on the block. If there isn't a suitable region, create a 'surface region' (Under Model:Features) beforehand.

By the way, a beam model of a space frame will not give a good estimate of the true stiffness of the frame, because beam joints are considered completely stiff, so your frame model will stiffer than the real thing. I quickly go from a beam model to a shell model of the frame, and I suggest you have a go at this method if you want good results. However, if time is short, go with the beam model but remember the real thing will be softer.

Regards,

Rod Giles Pro/Mechanica Users Group UK

P.S. Please feel free to get in touch, I especially interested in Formula SAE and helping out universities.

• posted

Thanks. This helps a lot. I will try some of these methods in the coming week.

Mike

>
• posted