Help with Fusion 360 designs

[DAPPH]

Hi . I've taken a series of pictures to show hte clearance of an inverted broken tooth between two still fixed teeth I can push a 20 thou feeler tip in so it cones out the other side ..suggesting each tooth is a tad smaller another picture indicates that as I've said the teeth sides are not machined verticals but have a tiny slope inwards to give a bit of meshing lead in when the parts are moving .

The picture with the rusty needles is a bit misleading for the broken body is more like a badly drawn 6 than a perfect circle .
See what you think .. had hoped for some sunshine for better pictures but had to give in and take them just inside the garage door whilst it's mizzling & very over cast .

Pictures Top 1
broken tooth between two good ones shows a gap .

2nd
A 20 thou feeler just slips through the gap.

3rd & 4th
Mic'd the tooth at it's widest point , note top of tooth to root is not quite vertical .

5th & 6 th
I can't quite get the narrowest width .. I thought the driving spline had sheared off clean at the base of the tooth , the measurement would be bout 10 thou less than indicated on photo 6 a micrometer reading of 0.186 "

7th & 8th
Final reading for wall thickness best I can get with a simple frame micrometer .

 

[Screwdriver]

Do you have easy access to the mating part? Might be easier to start from there.

 

[DAPPH]

Additional info
I checked the diametrical opposites with a steel rule and a few reasonably parallel things like... rusty needles and a precision drawn bit of spring stainless steel wire . it's difficult to say if they are crossed central about the axis as the collet frame is bust open & out of true but I suspect they are central
Se what you think .....

 

[DAPPH]

Do you have easy access to the mating part? Might be easier to start from there.

Yes but measuring it will be a nightmare

 

[Screwdriver]

Photoshop says definitely maybe.

The damage appears to overwhelm the required accuracy. What does the unbroken part look like.

PS a scanner would give you a better picture to tell from.

 

[Screwdriver]

If you scan both parts, each scan includes a steel rule, I might be able to help in photoshop. It's one of the ways I begin a digital model.

Oh and taking a mould off the mating part is probably a lot easier than making one up from scratch. The rest of the dimensions are trivially easy to reproduce.

 

[awemawson]

At the end of the day it's only a dog clutch - don't over think it. A bit of play is probably not going to be an issue, but not being able to engage it as it's too tight WILL be

 

[Screwdriver]

 

[DAPPH]

It looks like the sketch our daedakusminos Member has put up in post # 44

I like the reverse thinking .. wish I'd thought of that . So " Yes the lines all pass through the axis .

You can see the slight lead in on the tooth at 4o'clock and 6 o'clock if the were perfectly vertically machined they would be a sod to engage ..mind the lathe is about 77 yr old so they might be a smidgeon of wear on the cast metals but looking at the messing pressure points in the various photos it isn't much .


I'll have a think about a steel ruler & a scanning things . I don't want my darling giving me & an earful for several hours because I've got grease on her work scanner's glass .

Halfords electrical contact cleaner to the rescue followed by a few alcohol wipes ?
High level strategic talks will now ensue .

 

[daedalusminos]

Here's a model assuming they do cross in the middle (I've guessed waist height and depth)...

I've used 'User Parameters' so you can easily modify it (within reason)...

So it's incredibly easy to change the model by changing the user parameters...

and here's the timeline...

 

[Screwdriver]

From a mechanical perspective it makes sense for those faces to line up exactly though the centre of rotation. When you think about it, the castellations are actually two opposite engaging mechanisms (or levers). When drive is in one direction, all of the driven dog faces are on one side only, the opposite side is doing nothing. I should have drawn the guide lines in red and blue alternating colours but they are only there to see if the design is geometrically precise or if it actually does have a weird offset. Looks like simple geometry from here...

If you make the mould from the mating part, even a simple round casting of just that shape will give you perfect castellations and you can machine the cylinder to match.

 

[DAPPH]

Here's a model assuming they do cross in the middle (I've guessed waist height and depth)...
View attachment 245105
I've used 'User Parameters' so you can easily modify it (within reason)...
View attachment 245108

So it's incredibly easy to change the model by changing the user parameters...
View attachment 245111

and here's the timeline...
View attachment 245115

Wow , would you send me the file so I can print it or slice it for a true Prusa i3 Mk 3 printer I will PM you my email in anticipation
Thanks
Dave .

 

[daedalusminos]

No problem, I've added the tooth draft angle and height to the User Parameters, and the most important parameter 'OuterMajorGapWidth' defines this dimension:

 

[daedalusminos]

Prusa i3 Mk 3

How are you finding the MK3? After years of using a prety generic perspex frame Prusa clone (which has paid for itself many times over) I've recently ordered a MK3s - just waiting delivery.

 

[Screwdriver]

I'm still wondering if a plastic part itself might be sufficient for this application. Crossfeed shouldn't require too much torque, if it does, you're probably pushing the machine too hard.

ABS would be the better candidate or perhaps PETG? Once you have a model you've got nothing to lose by trying it...

Maybe someone here could even figure out the relative stresses.

 

[daedalusminos]

I'm still wondering if a plastic part itself might be sufficient for this application

A high glass fill plastic might hold up, not so sure about anything 3D printed - obviuosly something cause dthe original metal part to fail.

 

[Screwdriver]

A high glass fill plastic might hold up, not so sure about anything 3D printed - obviuosly something cause dthe original metal part to fail.

Some thing or some one?

 

[Screwdriver]

In all seriousness though, for that to work you'd probably also have to print it on it's side for the layers to orient themselves for maximum strength.

 

[bigegg]

In all seriousness though, for that to work you'd probably also have to print it on it's side for the layers to orient themselves for maximum strength.

Try printing (or boring after) 3mm holes through each tooth, and epoxy in a piece of wire as reinforcement?

 

[slim_boy_fat]

Files are so last year .
I have made a small vertical milling vice from a small precision round vice using precision ground key way steel as the mounting bar . It fits my QCTP tool holder so I could in theory mill parallel teeth to the smallest dimension using the 120 toothed saw blade on a collet to get initial indexed accuracy . I suppose a with homemade micro adjuster & a precision turned down parallel 1/4 inch router bit cutter in the four jaw chuck I could also get the right taper required .. Something to sleep on for a few nights perhaps .

Can you let us have a pic, I'm trying to visualise your setup but it's getting late and my head's full of fog.