Belt/chain to linearly drive carriage

[arsolidworks]

Dear all,

I am currently putting in place a linear guide. This assembly is composed of 2 pulleys, a belt (showing in red on the joined photo) and a sliding plate.

Currently I can create the belt/chain system between the 2 pulleys (excluding the sliding plate): this works fine and when I rotate one pulley by hand, I can see the other one rotating accordingly.

Now I am actually expecting the belt to drive the sliding plate. i.e. I am expecting to link the plate to the belt, the belt fitting a groove machined in the plate (showing in yellow on the joined photo). The belt is supposed to be tightened to the plate by a part not shown on the joined photo).

The problem is that the belt/chain feature does not seem to accept flat surfaces as input. I was actually expecting the belt/chain feature to cope with flat surfaces in order to have:
* the belt deform to account for a misalignement between the belt and the groove,
* the plate driven by pulley rotation (to transmit power from the belt to the plate).

I thought my design was really common and would be handled by SW out of the box. Did I miss something?

Do you have any idea how to design such a system in SW?

I have joined the assembly and part files of the exemple for a clearer view.

Regards,

Antoine.

 

[arsolidworks]

I do not seem to be able to insert images, so here it is attached:

 

[SincoTC]

Hi arsolidworks

I created something similar a while back and found that the only way to make the belt "drive" the carriage was to create an additional Rack and Pinion Mate using an edge of the carriage parallel to the travel and one of the pulleys as Mate reference selections and then specifying Rack travel/revolution.

I hope this helps.

Trevor Clarke. (R & D) Scientific Instruments.Somerset. UK

SW2010x64 SP4.0 Intel Core i7 2.94Ghz, 12Gb Ram, NVIDIA Quadro FX3700 Driver: 6.14.11.9100
SW2010x64 SP4.0 Intel Core 2 Duo 3.17Ghz, 8Gb Ram NVIDIA Quadro FX3700 Driver: 6.14.11.9100

 

[arsolidworks]

Hello Trevor,

This solution does not check all the boxes (as the length of the belt is not computed accordingly) but this is a nice workaround which I had not thought about. I will probably end up using it.

Thanks a lot!

Antoine.

 

[SincoTC]

Hello Antoine,

I assume that the belt length is not computed correctly because it's not pulled against the face of the slot you want to clamp it into and the Belt mate won't allow the face of the slot to control it? If so, it may be possible to fool it by positioning a couple of "dummy" idler tension rollers of minimal diameter on the carriage, acting on the back of the belt and placed so that the belt is deflected into the desired position. this should give a snapshot measure of the belt length, but not sure without trying it if it would break the mate if the carriage was moved.

Good luck!!



Trevor Clarke. (R & D) Scientific Instruments.Somerset. UK

SW2013x64 SP2.0 Intel Core i7 2.94Ghz, 12Gb Ram, NVIDIA Quadro FX3700 Driver: 6.14.11.9100
SW2013x64 SP2.0 Intel Core 2 Duo 3.17Ghz, 8Gb Ram NVIDIA Quadro FX3700 Driver: 6.14.11.9100

 

[arsolidworks]

Hey, this is another great workaround! I guess the length will change over time which makes things a big complex, but this will definitely givr a finer estimation of the belt's length!

Thanks Trevor, your help has really pushed me forward!

Antoine.