TurboCAD Mac Forum

New TurboCAD Mac 2D/3D Training Essentials

Modeling wave washers
Read 453 times
* June 28, 2018, 09:02:01 PM
I've been having fun/frustration trying to model wave washers - aka wave springs, the things often used for preloading bearings. 

It's fairly easy to generate a 4-lobe washer:

1) Draw circles centered on the origin for the ID and OD and extrude them by the washer thickness.

2) Draw center-point reference lines from the origin at 0, 45, 90, and 135 which extend beyond the OD of the washer.

3) Center-bend the washer up by a small amount (e.g. radius 0.5", angle 8) along the 0 and 90 lines, and bend it down by the same amount along the 45 and 135 lines.  The endpoints of the reference lines are used to define the bend axes.

The only problem with this is that all of the wave washers I've got only have three lobes, not four!  I did manage to model a 3-lobe washer that's kind of close to correct, but it was a challenge to figure out:

A) Create a 180 (+90 to -90) segment of the washer from two arcs and two "end cap" lines.

B) Add + and - bends at 30 and -30, and then add smaller reverse bends at +/-60 and tweak these angles until the 60-90 ends lay flat in the plane.  Using my 8 primary bends, I needed 4.6 secondary bends to get this close enough.

C) Make a pie-with-a-120-bite-missing cylinder from -60 to +60 and subtract this from the washer segment to leave a 120 piece. 

D) Polar duplicate the 120 segment to create two more 120 segments.

E) Do two solid unions to join the three pieces together together.

The result isn't an accurate representation of a 3-lobe washer - it should have positive bends at 0, 120 and 240 and negative bends at 60, 180 and 300.  But I've spent two hours getting this far, and it's after midnight...

So my question is: is there a better or easier way to do this?

A screen shot of the 3- and 4-lobe washers, and the .tcp file of what I did, are attached.

>>Computers were invented to help people waste more time faster<<

June 28, 2018, 11:34:23 PM
Hi Kihillig! Welcome to the party!
I see your resolution is a touch low- go superfine, and the same with lines,and as simple as this is, is " difficullt"! This is one of those times that we peacrice what a CAD guy told me when I was starting out! :
You also have a join which is why you need a higher resolution, and prolly all 8 digits for accuracy!
I think try a spline with the control points @ the required locations, + /_ above the zero line( circle) , with extra for the distance for the "true position" I have a long day tommorow, so this is as far as I can go for now.
One Question : Is this for "real"?= a True part that will be formed by a machining source, or "for fun" (RC birds, 3D Printers etc?


* June 29, 2018, 01:24:04 AM
1. Draw a center point circle, on this circle use the points along curve tool and enter 8 points, six for 3 pt. Select circle and Edit -> Change Object Type to Interpolate spline. Then "Show Points"
2. Select the "Add Spline Point" tool and add a spline point at each point located by the "Points Along Curve" tool. Use the "Remove Spline point" tool to remove all the other points. Delete the points placed by the "Points Along Curve" tool, easier to select control points.
3. In isometric view enable the Gripper. Select the "Select Deep" tool and holding the shift key select the four quadrant control points. For the 3pt select the 3 pts at 120 and using the gripper, move the points up along the Z-axis. Deselect all and select the spline, hold the option key and with the gripper scale the spline making a second spline.
4. Select the "Skin Surface" tool and click on each spline creating the surface. Thicken using the "Thicken Solid" tool.




* June 29, 2018, 06:57:41 AM
Thanks, Mitch!  So far I've been doing all my modeling using lines and solids so wasn't aware of the tricks one can play with surfaces - and I hadn't run into anything needing spline curves yet.

The "splined" circle appears to be slightly different from a "true" circle, but not by enough to matter in this case.  I did try to sweep a rectangle (i.e. the washer cross section) along the spline curve but this just gave me a "curves not planar" error - which seems odd, since I can sweep along a helix or other non-planar curve...

Johnny - this is a real part, but not one that I'm going to make.  It's just something I need for an assembly drawing, and I'm obsessive enough to want it to look right.  The version I came up with has some kinks due to the secondary bends - it's not a resolution problem, it's just a less-than-perfect model. 

>>Computers were invented to help people waste more time faster<<

June 29, 2018, 09:29:37 AM
Kool! I like obssesive! nuthin' loke the real deal! I'm glad Mitch got ya along farther than me! :D


* June 29, 2018, 10:43:23 AM
I saw this as a surface/organic modeling problem.

The "splined" circle appears to be slightly different from a "true" circle

Yes, the more points the truer the circle but the downside is dealing with all the points.

I did try to sweep a rectangle (i.e. the washer cross-section) along the spline curve but this just gave me a "curves not planar"
You may have needed to set the workplane to side if you were in a top view. Most of the extrudes require planar geometry but it's not required for the path.

The easiest way to do this is to go to the Mcmaster-carr site and download the needed cad file.