Generating inner rotor profile of a Gerotor pump based on a law curve in NX 1

[wkucinski]

Can anyone inform me on how to properly define my expressions in NX.

I've created my profile in an Excel Spreadsheet to validate my equations but I cannot get the law curve function in NX to generate the profile.

The gerotor inner profile is generated by the offset from a trochoid.

My expression window is attached as well as an image of the two profiles created in Excel with the same equations.

 

[wkucinski]

expressions

 

[cowski]

I'm not sure that this is the problem, but it may play a part.
In some of your expressions, it looks like you want to use degrees as the angular measurment (e.g. tt = t*360), but the expression is defined using the radians angular measure.

www.nxjournaling.com

 

[SDETERS]

What we did was export the different points from the Excel sheet. Then we generated a spline through these points in NX. Be interested if there is an answer to this.

 

[TaylorAnderson]

Can you post your spreadsheet for reference?

I'm pretty sure this is just a degrees/radians dance, and having a functional spreadsheet would help unravel the mystery.

Thanks!

Taylor Anderson
NX Product Manager, Knowledge Reuse and NX Design
Product Engineering Software
Siemens Product Lifecycle Management Software Inc.
(Phoenix, Arizona)

 

[TaylorAnderson]

So, I kinda started over here...

I was struggling to understand what your expressions were doing (I just didn't understand what all of the constants were indicating, in the end) and so I went back to your textual description of this curve as "an offset from a trochoid" and started with the definition of a trochoid (an epitrochoid on the outside of a circle, as opposed to a hypotrochoid around the inside of a circle) mostly just because that's what your Excel graph picture looked like.

And then I built that curve as a law curve, and added an offset curve feature after it:

We have the modeling tools, after all... Why not use them?

It turned out quite nicely:

The expressions to set this up were as follows:

(The image is from NX 11, but the math will be identical in earlier versions.)

In order to maintain a closed loop, [big] R must be a multiple of [little] r, and so the "multiple" expression above serves to enforce this condition.

And just because I can, I threw a Product Template interface on this model using PTS (Product Template Studio) to easily control all of the parameters:

And so by varying the "multiple" expression (labeled "Number of Lobes" in the dialog) we can easily generate the various shapes in this family:

(Number of Lobes = 5)

(Number of Lobes = 4)

(Number of Lobes = 3)

One thing I noticed... The parameter/expression "d" (distance from the center of the smaller "rolling" circle to the point being traced) always needs to be smaller than the expression "R2" (the radius of the "rolling" circle) in order for the offset curve feature to update correctly. You can see the conflict start to develop as d approaches R2 here:

(R2 = 5.50 and d = 3.00)

(R2 = 5.50 and d = 4.50)

(R2 = 5.50 and d = 5.00)

(R2 = 5.50 and d = 5.25)

If the epitrochoid ever gets into the "full point" mode (where d=r):

(R2 = 5.50 and d = 5.50. The HD3D tag and the warning are from my Product Template interface.)

or into the "loopy" case (where d>r):

(R2 = 5.50 and d = 8.00)

...then the offset feature will fail, because there are some gnarly indeterminate offsets at the pointy cusps.

Does this help? Is seeing how I've set this up useful at all?

I suspect that if you've got an equation for the combined trochoid and offset, that we could get it working here in one shot. But I'd want to see the full equation and understand the constants a bit better.

At any rate, thanks for the fun investigation. :-D


Taylor Anderson
NX Product Manager, Knowledge Reuse and NX Design
Product Engineering Software
Siemens Product Lifecycle Management Software Inc.
(Phoenix, Arizona)