How to optimize press fits?

[tharding247365]

We press 2 round shafts together. One piece stainless, the other ceramic. We require a specific pull force, and as of now we sometimes get good parts and sometimes bad parts. Parts are not large, they're about half an inch big.

Are there are things to consider when trying to achieve a consistent pull force or to increase existing pull forces?

Knurling or other factors or means of mechanically fastening these 2 parts together without doing anything exotic?Production does not want to deal with adhesives.

 

[hendersdc]

Can you provide pictures of the parts and a description of how they are used, how much force they need to resist, and why a press fit was chosen?

The first thing that comes to mind is why aren't threads used to hold the parts together?

 

[dgallup]

Press fits are fine for many applications but you need to control size, roundness, surface finish, hardness, entrance condition, etc. very closely to get consistent reaults. We had an annealed part that was giving us fits pressing it on to a hard shaft. Turned out the anneal was leaving a thin oxide film on the parts that was like microscopic sand paper. When we changed to a bright anneal the problem went away.

Same material and hardness is asking for problems as you are likely to get galling.

More details will get better answers.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[tharding247365]

Left out the most important detail. One part is ceramic, other is stainless. Hence why no threads.

 

[MintJulep]

Because it's the interference that matters.

Big shaft into small hole different from small shaft into big hole.

If you want things more consistent you need to either:
Reduce the variations in both parts.
Measure and match to control the interference.

 

[chicopee]

With ceramic isn't there a very good chance of cracking it with too much interference fit or better known as clearance fit? If the ceramic shaft is to fit inside the stainless steel shaft, heat up the SS shaft for the ceramic shaft insertion and let it cool. It seems that some statistical testing is in order following a clearance study.

 

[Tmoose]

Was the "pull force" chosen just as a method to "inspect" the quality of the press fit?
You say there is a "specific pull force." I'm thinking ( GUESSING !! ) you really mean there is a minimum, but would like to hear the real details.

When establishing the "specific pull force", was "it" determined empirically by testing components assembled with the full range of acceptable fits ?
Are the "bad parts" inspected to determine what their actual assembled fits were.

Or is the "specific pull force" related to the operating environment ?

A QA test that flunks parts that (1) are made to the drawings, and when assembled (2) satisfy all the requirements for form, fit and function
is a terrible test.

As mentioned by others, //perhaps// selective fitting will be required if parts made to the drawings do not always fit and function.

 

[tharding247365]

Parts are inspected when they come in. The pull force is indeed a minimum that we have call out. There was testing of acceptable fits before my time. Pull forces are not done every part, but as sample quantities to verify press fit

I'm looking at possibly heating up the stainless part, and pressing in after. Hopefully this really grips the ceramic when it cools.

I'll trial out how much interference I can get before the ceramic starts to crack, and then of course back off and test the pull forces along the way.

Any other suggestions? Will be a bit before I try this.

 

[tygerdawg]

This seems like it would be a good opportunity to employ a six sigma / process improvement person to investigate this using Design Of Experiments methods. Of course there are "standards" for press fits. But your situation seems to need deeper investigation into the various process parameters, features, materials, finishes, and so on. There may very well be interactions between all the variables that are not obvious that cause poor quality & erratic performance. Undergoing a DOE study may provide definitive work piece dimensions and features and guide you to an improved means of measuring success.

TygerDawg
Blue Technik LLC
Virtuoso Robotics Engineering

http://www.bluetechnik.com

 

[racookpe1978]

Is the stainless the internal part, or the external part, of the two-part assembly?

The ceramic will not withstand "any" expansion forces, and "might" withstand a minor bit of compressive force (think of a metal part crushing the inner ceramic liner, but the crushing is not sufficient to turn all the ceramic to sand.) However, any crushing (or cracking as the internal metal expands against the surrounding ceramic liner) will cause the two-part assembly to fail.

Which direction is the force failing? Push to right end, or push to left end, or both equally, or is it a turning moment? You might be able to fabricate a lip or two expansion clips at both ends of the ceramic if the largest force will be from one end.

 

[MintJulep]

If your parts are "within tolerance" but your assemblies don't work then your tolerance are wrong for the process used or the materials.

 

[hendersdc]

Still guessing since you haven't shared any pictures or diagrams but what about changing the design to use retaining rings, cotter pins, stepped shafts, bearing nut, or anything else that would be easier to assemble?

 

[tharding247365]

Appreciate the feedback so far. Can't share any pictures.

Racookpe1978: The stainless part is on the outer side of the press fit assembly, and the ceramic solid shaft is the inner part.

TygerDawg: It does look like somebody will have to look into Design of Experiments. Measuring parts in an excel sheet with pull force values to get a baseline, and do further investigation on the press fit call outs, with additional test fits, and other variables.

MintJulep: I'm aware of this. Hence why I came here for advice. Looking for other tips and tricks on a very non-standard press fit application.

Hendersdc: We can't really deviate from the current design because of the industry we're in. Looking to optimize the current design the best we can. I would do a complete redesign if that was on the table.

I believe the original design was smaller, and was scaled up for this application. Maybe looking at a longer press fit will add more friction if that makes sense. Then also look at heating up the stainless part, and letting it shrink around the ceramic shaft if nothing else wants to stay consistent.

The sad part is, the loctite adhesive we've tested as worked wonders, but the production floor doesn't want to use it for some reason.

 

[Stick.]

If the Loctite adhesive truly works wonders, you should figure out why the production floor doesn't want to use it and then consider removing that roadblock (depending on what it is).

 

[MintJulep]

Correct approach:
1. Understand the cause of the problem.
2. Solve the problem.

The path you appear to be heading down:
1. Guess and hope for the best.
2. Repeat.

Plenty of good advice on this thread to get you started on the correct approach.

 

[Tmoose]

When it is ascertained the assembly function is OK with the largest bore and smallest shaft ( per drawing ) then the test needs to be optimized by testing several assembled pairs with that minimum interference.

But as others mentioned, if a couple of $50 bottles of Loctite ( one for each assembly station) make all assemblies "pass the test."
there is unlikley to be a less expensive "fix."

https://www.grainger.com/content/general-catalog?pagelabel=2222&search=4KL51

 

[rb1957]

You could assume that the failures results from the worst adverse tolerances (producing a minimum interference).

Why not tighten tolerances (restrict the smallest shaft and the biggest hole) ?

You're inspecting the assembly … do you verify that the pieces of the failures are within tolerance ?

another day in paradise, or is paradise one day closer ?