Delrin Set-Screw Fastening

[USMechE6]

Hi All,

We recently had a steel set-screw (supposedly Loctited) back out on a moving part of one of our critical machines, causing a whole lot of damage and expense. Talking about a $1 item damaging a multi-million dollar piece of equipment. One thing I am considering moving forward is to put a delrin set-screw behind it. This will enable it to act not only as a jam-set-screw, but be much less damaging to equipment if it did back out.

I've experimented with different grades of Loctite and screw-head profiles that make it so that the set-screw will just barely loosen with a tool without stripping (what I want). This is at about 120 degrees F. Given that the tapped hole is steel and that delrin expands so much more than steel with temperature, should I expect that to minimize risk of loosening to the frictional force of expanding against the steel tapped hole, or just cause it to back out based on the path of least resistance. I would expect the latter more likely if I wasn't using Loctite, too.

I realize this is more of a philosophical question that may be answered through some form of calculation, but I'm not sure what that would be exactly given all the variables, so I just wanted to throw the idea out there and get thoughts.

Throwing in the registered trademark of Delrin from DuPont and Loctite from Henkel so I don't get in trouble

 

[drawoh]

USMechE6,

I have learned to not trust set screws. A set screw digs into whatever it is clamping and plastically deforms the material. If you apply significant torque to it, the material deforms more and everything becomes loose.

Delrin is very resistant to chemicals, including thread-lockers, and it has a low coefficient of friction. Delrin does all sorts of cool things, but clamping is not one of them. I have seen very cheap, shoddily cast screws with the molds out of register. These actually are self-locking. You will have fun telling your vendor that you need the worst crap he has got. It still will not exert much friction on your steel set screw.

I cannot think of a shaft clamping system that is inferior to set screws. Here is one better solution.



--
JHG

 

[MintJulep]

As a material for screws Delrin is pretty close to useless. Delrin is plastic. Plastic deforms plastically at low loads relative to not plastic things.

I agree with drawoh that just about every other way to attach something to a shaft is better than a set screw.

 

[Nescius]

How often does the set screw need to be backed out of the hole? Does it ever need to be adjusted?

Crude suggestions follow:

If the head of the set screw is below flush, a heavy stake at the the edge of the hole would prevent the screw from coming out. Think very heavy center punch blow, driving material inward to violate the hole boundary. Set screw could be loosened, but not removed. Grind away staking to remove set screw.

If the head of the set screw is flush, grind one or more reliefs in the head of the set screw to receive the staked material...not adjustable.

Definitely not sophisticated, and maybe not at all what you're looking for, but worth a look.

Also, you may believe you want the screw to barely be removable before the head strips out...and if that's how it behaves, you'll always be happy. The caveat is that this regime is right next door to actually stripped out.

 

[3DDave]

If it backed out I think it did not have loctite. Maybe someone used anti-galling materials instead or just used grease.

Most often the screw does not back out, the end gets torn up or the shaft gouges out. People think it has come loose because the screw moved and they are usually wrong.

More details = better answers.

 

[Jboggs]

I hate set screws and avoid them if at all possible. Any time you use a set screw in a design you must ask yourself what's going to happen when it fails, because it WILL fail. Its not IF it fails, its WHEN. Find a way to redesign whatever components are affected to eliminate the set screw and use some other method to accomplish its task. If this is a shaft and hub assembly there are numerous keyless hubs that work very well. If its just clamping a flat surface there are more reliable ways to do it. In my opinion, set screws are evidence of sloppy or immature design.

 

[JNieman]

Jboggs' post seemed extreme and I wanted to compose a retort bringing up valid use of set screws.

I couldn't.

The only time I've used set screws were for one-and-done tools that would be disposed of as soon as they served their short-life purpose (deliberately sloppy design)
or as crude elevation screws to balance/align something in-situ but that's rare and not suitable for long-life (immature design)

I guess I agree. Set screws suck.

 

[Screwman1]

I agree that set screws are not a good design, but if you are stuck with them, then you have to work with what you've got.

I think that your delrin concept is flawed because of the delrin material, it's shear strength is so low as to be virtually trivial compared to any outside forces acting on it.
I would guess that your original failure was due to either a lack of Loctite or a lack of cleaning prior to using the Loctite. For Loctite to really adhere well you need a clean and dry surface. You need to flush both the set screw and the tapped threads with a strong degreaser that will leave no residue. Then you can add the Loctite to the full OD of the set screw. If you really want to make sure that the set screw does not come out, use the Loctite 270. This will NOT come out unless you heat the nut member to about 300 degrees C.

 

[KENAT]

So, unless I'm missing something no where does the OP give the explicit application of said set screw.

There are situations where a set screw is a cost effective adequate device.

Clearly in this case, if the consequence of the set screw failing is so significant then this may not have been such a situation.

What are the materials of the screw and threaded hole? Loctite thread lock only works properly with active surface chemistry from what they told me, for other materials such as SST or AL a primer will probably be required for it to work properly. (Yes, I've seen SST fasteners threadlocked into aluminum parts etc. that wouldn't come out, might have been galling or maybe just 'lucky' I don't know.)

For a simple fix how about going to a patched screw? Similar idea to nyloc but on the screw. This way someone can't forget to properly apply loctite - unless they install the wrong type of screw which is possible.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[USMechE6]

Thanks all for the responses thus far. I definitely agree that no doubt set-screws are unfavorable for all of the posted reasons. I should have explained the application.

This is being used in a steel piston assembly, holding a steel pin in that it inserted through the slot of a valve that's on the end of the piston. The pin is perpendicular to the motion of the piston. As the valves opens and closes its motion is restrained by the interaction between the pin and the slot. The steel set-screw essentially caps the hole used to insert the pin from the outside of the piston body, keeping the pin in. It is necessary for it to be removable for maintenance and the equipment cannot be heated to 300 C let alone above about 70 C.

We've used plastic set-screws in the past, but they've often stripped to the point where it had to be drilled out, opening the piston hole from 3/8"-16 to the current 7/16"-20. This was because we really hadn't experimented to determine the optimal combination of Loctite and head profile. Doing some (rudimentary) bench-top tests has yielded the current combination I am considering. We did try using the recommended Loctite primer, but that ended up making the adhesion too good and would cause the screw to strip upon removal.

There are better solutions in general, but for my specific application I am kind of constrained. Perhaps a spring-loaded detent perpendicular to the tapped hole may work.

 

[btrueblood]

Would a roll pin work in place of the pin, or do you need to seal the pin hole against leakage?

 

[USMechE6]

A roll pin might work except for the fact that it needs to also go through another hole on the other side to ensure that it's fully through the slot. I suppose I could taper that other hole for ease of placement/location. Only thing is that the pin is removed by tapping it out from that other side through a smaller diameter opening that hole connects to but prevents the pin from falling out there. Being hollow would make that difficult, unless I had it taper down slightly to a closed/flat end.

I'm still not sure about the answer to my original 'philosophical' question, but by no means mind the digression to alternative ideas - very helpful!