Dynamic Mechanism Bearing Reaction Forces

[JonReynolds]

Hi folks,<BR itxtNodeId="117"><BR itxtNodeId="116">I am trying to work out how to get bearing reaction forces within a rotating assembly.<BR itxtNodeId="115"><BR itxtNodeId="114">Picture a simple shaft, with an eccentric mass towards one end (kinda 'L' shaped), then support it in 2 bearings at each end and rotate it. <BR itxtNodeId="113"><BR itxtNodeId="112">The eccentric mass is closer to one bearing than the other, so I would expect the bearings to see different loads.<BR itxtNodeId="111"><BR itxtNodeId="110">How do I determine these loads in Pro/Engineer?<BR itxtNodeId="109"><BR itxtNodeId="108">I can set up a mechanism to rotate the assembly, but I am not sure how to constrain the shaft in 2 different places, i.e. at each end in seperate bearings, so that I can measure reactions in each bearing.<BR itxtNodeId="107"><BR itxtNodeId="106">Any help much appreciated.<BR itxtNodeId="105"><BR itxtNodeId="104">Thanks in advance.

 

[Kaz Z06]

Use a bearing joint to support one of your bearing positions and use a 6DOF joint at the second, while constraining all 3 translation DOFs for the 6DOF. You'll need to construct a "radial" reaction for the 6DOF joint from the X & Y components. The bearing joint has a radial reaction load as a built-in measure.

 

[JonReynolds]

Hi Kaz,


Thanks for your reply. Could you please explain a little how to use the 6DOF connection? I have never used it before.


Thanks again

 

[Kaz Z06]

Jon,


6DOFS are easy to apply - you need a co-ord system on each part at the locations they are to assemble (and of course oriented the same way). You also need a datum point at the center of each co-ord for each part. You use the datum points for "zero" references on all translational DOFs. There should be enough help on-line to guide you through it.


Chris

 

[JonReynolds]

Hi Kaz,


Please can you elaborate a bit more?? I am sorry but I just can't seem to Google for the right things... no one seems to know about this black art of Mechanism Dynamics...even Pro/E doesn't seem to be able to tell you in the Help menu.


I have tried adding points and CSs on my shaft and 2 bearings. I have tried a bearing connection at one end and a 6DOF at the other, but i cannot get it to be tied down so that only rotation is allowed.


Trying to set max and min limits to 0 on the connection translations is not working.


Please help!!


Thanks





p.s. this forum has a NASTY habit of losing your whole post if you type the Captcha code in wrong. That is VERY BAD.

 

[JonReynolds]

Ok, Think I am getting somewhere.


Picture a simple shaft with a bearing at each end. Centre of each bearing has a CS and point. There are CS and points on the shaft that line up with the bearings.


I used 6DOF connections between the CS on one bearing/end of shaft and the same at the other end, using the points to reference for translation.


In Mechanism, I then added a servo motor one each translation axis, so 6 in total (2 bearings) all with a position of 0. EXCEPT for one, which I put in a velocity, i.e. my driving motor.


This produces the desired mechanism, i.e. it turns as expected.


I can now graph connection reactions on each connection to get radial force at each bearing.


My only problem is that the numbers I am getting are not what i am expecting when I calculate the bearing forces by hand. Is there anything in my process that I am missing?


Many thanks!

 

[JonReynolds]

The numbers I get are in the order 50% smaller than expected.


Substituting the 6DOF for a cylindrical connection and measuring the connection reaction on that I get the right numbers.


So the question is now why do my 6DOF connection reaction give different numbers?

 

[JonReynolds]

OK, got there in the end

 

[Kaz Z06]

Did you vectorally add the X & Y loads from the 6DOF to get the total bearing load? You must do this to compare it to the "radial" load the cylindrical joint will provide "out of the box".

 

[JonReynolds]

There is a "Total Force" as one of the components for the 6DOF connection reaction measure, which is the resultant of the 2 XY vectors.


The numbers are all coming out as I expect from my hand calcs so I believe all is good.


Thanks for your help!

 

[JonReynolds]

On the 6dof joint, I can get vertical/horizontal and total load measures.


Do you know how I could also get vertical/horizontal measures out of the bearing connection end?


Thanks.

 

[Roger]

I never cease to be amazed at how much time is spent trying to get a computer to do what I was taught to do MANUALLYin JUNIOR COLLEGE!!! Just how precise must your bearing calculations be? Are you going to Mars? Need I remind you that we went to the Moon using a SLIDE RULE!


I'm sorry folks, but, "degreed" engineers put me down because I only have an AAS degree. It's really funny how I can do things manually that "degreed" engineers need a computer to figure out.

 

[JonReynolds]

Well I hope you've made yourself feel better with that extremely unhelpful
post.

I am perfectly capable of calculating the bearing reactions of a simple
rotating shaft like in my example. I am perfectly capable of calculating,
by hand, the reactions of multiple bearings along the length of a multi-
throw crankshaft and then further to determine stresses and fatigue
factors of safety, not to mention torsional vibrations.

The trouble with hand calculations though, are they are only as good as
your inputs. When you have a complicated shape such as a cast
crankshaft, with ( almost organic) undercuts and pockets, the inputs
required to calculate the reactions along the length become impossible
to estimate. One ends up using the tools available in the CAD package
toacquire the necessary inputs.

One can then either take these inputs and calculate things by hand or
one can make use of the tools available and perform the calculations
directly in the CAD allowing for a much faster and accurate answer and a
more efficient use of time . Plus the ability to optimise the design by
setting goals.

That's how we design things in the 21st century. Hand calcs are
important, but are the starting point.

 

[JonReynolds]

Ok, well through trial and error, I think I found a solution.


I used the following connections on my rotating shaft:


-A point (on shaft bearing centre) to surface (in assy inline with one bearing) - this controls translation along rotation axis.


-Two 6DOF connections on CSs located at each bearing centre.


Then in Mechanism, servo motors on each vert/horz translation axis of the 6DOFs to stop the rotating shaft deviating from rotational centreline plus a servo on one of the 6DOF connections rotational axes to drive the mechanism.


Measures can then be placed on each 6DOF connection in both vert and horz directions to get the different reaction at each bearing.


Which can THEN be verified by hand calcs.

 

[Kaz Z06]

Jon,


This approach will work well. Bearing joints don't have the ability to resolve loads in radial directions - just the total radial load. Using a 6DOF will allow you to measure loads/momnets in all directions after properly constraining the DOF that need to be "locked". You can use the 6DOF to model bearing stiffness to some degree too with springs acting in the X-Y directions instead of a servo motor that fixes the DOF.