Thread Strength Of Molded Plastics

[dapeake]

Greetings,

I am seeking information on the strength of threads in various plastics.

The image below shows my part. It shows a threaded zone: 5/16-18, the length of which is .377". It needs to withstand repeated, non-shock (gradually applied) loads in excess of 100 lbs in both directions.

Note: I can lengthen the threaded zone, or switch materials, but without data on the properties of the materials, I am shooting in the dark.

Any help in nailing down the thread pull strength of various molded plastics would be greatly appreciated.

Thanks,

Dan

 

[patprimmer]

For an unfilled grade, Acetal homopolymer is the best at a reasonable price.

Ticonaor Dupont or Asahi should have actual data. The calculations should be possible from quoted shear strength and thread profile.

Plastic threads are very dependent on correct design for plastic. Traditional metal thread profiles are far from optimum for plastic.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

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

for site rules

 

[Pud]

Use the correct screws:

e.g.

http://www.prelok-fasteners.com/fastenerspt.html

Some sums here:

http://www.aspltd.co.uk/pt-technical-info.pdf

And here - scroll down for Screws Design Software.

Harry

http://www.tynevalleyplastics.co.uk

 

[TVP]

Self-tapping fasteners like the Delta PT are recommended for engineering thermoplastics like nylon (PA) and acetal (POM). Here are a couple of design guides from Ticona and DuPont with some information on thread strength:

http://www.hipolymers.com.ar/pdfs/celcon/diseno/Designing with Celcon acetal copolymer.pdf

http://plastics.dupont.com/plastics/pdflit/europe/design/L12565_9.pdf

There is a considerable amount of information available in technical papers from SAE and SPE (ANTEC conference papers). Search their websites for specific papers.

 

[telecomguy]

You mention 100lb load "in both directions". From the geometry, it appears you should consider fatigue due to non-axial loads as well.

jmho,
-tg

 

[lumenharold]

I had been assigned a project to improve screw pull-out performance from plastic bosses in parts made from polypro and styrene. A couple things that were very important in addition to the screw selection and boss design issues already listed that I would suggest are: Work with a good tooling person. Getting consistant draft and diamter size in the pilot hole is critical (I was dealing with blind holes so it was a little different than your illustration). You can calculate stress all you want but if the hole is shaped like an hour glass or a pear it won't perform like you expect it to. Second, get access to a pull tester and correlate the data to an existing design. That way you'll know what to expect with a little confidence.



Harold
SW2009 SP2.0 OPW2009 SP0 Win XP Pro 2002 SP3
Dell 690, Xeon 5160 @3.00GHz, 3.25GB RAM
nVidia Quadro FX4600

http://www.lumenflow.com

 

[Pud]

Could you mould in a threaded metal insert?

Unless you are using unscrewing cores to mould thread, it may also be cheaper than post-tapping (which is not a good idea, anyway) Here's a UK based supplier:

http://www.tappex.co.uk/index.aspx

part to illustrate.

With whom I have no connection at all.

Harry

http://www.tynevalleyplastics.co.uk

 

[rb1957]

or is it plastic-in-plastic ?

if it is plastic-in-metal, what does it take to strip the threads ?

 

[dapeake]

Thanks for the replies!

I left quite a few unanswered questions in my original post. Thanks for bringing those to my attention.

Function: The end device (product) is meant to install and extract a bearing from a from a plastic housing, by jacking with a metal, hand turned, 5/16-18 screw. So the load is virtually 100% axial. The part will be subject to repeated daily use.

Our plans start to fall apart as part price crosses $.30USD (lot size 5kmin, to 50k, 5 orders per year) If molded, I have a strong desire to keep this produced in the US (my molder, whom I have used for about ten years, just had twins!)

Harry, yes, molding-in a threaded insert is an attractive way to go. I am, however, concerned about piece price. I would be very curious to learn the pricing and form factors available.

Options to consider:
-Tap into molded plastic or die cast as a 2nd op
-Mold in threads (die case or plastic)
-Threaded insert in plastic
-Screw machine
-Mold this in two pieces so they press-fit together capturing a regular nut inside.
-Mold the part with a hex pocket on each side. Leave nut on the jacking screw (to reverse function -install/extract- flip housing, seating nut on new side) drawback - possible loss of nut.

Do any of these approaches stand out as a superior cost saver?

To tg: thanks for pointing out the non-axial weakness; unseen in the above graphic are ribs in the shelled annulus.

Here is an image of the most basic, utilitarian form of the part (no draft, no shelling, i.e. how a screw machine might make it.)

Thanks also to those who went to the effort of posting links - good stuff!

Thanks,

Dan

 

[dinjin]

Why isn't it threaded thru the whole part and not have the counterbored 5/16 holes top and bottom? If these are jacking threads, I would think they would have to be at least 2.5 times the thread diameter in length minimum.

 

[patprimmer]

It will need to be a metal thread. In my opinion plastic threads will fail unless the required pull force is extremely low.

The cheapest should be a hexagonal recess for a nut.

One better would be to retain the nut with a snap fit collar of some sort or maybe a bayonet type fitting. This obviously should be on the side that is not loaded in use.

The best solution is as Harry suggested to mould in a threaded insert.

Another similar alternative is to instal the inserts with ultrasonic vibration.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

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

for site rules

 

[dapeake]

Thanks Dinjin,

True, the .377 section is a bit short for what I am doing. Acceptable for metal, but probably unacceptable for plastic.

To address your question; overall part length is a concern for us, and the jacking screw has features on it which can not go into the threaded area. Jacking from both sides of the part; considering the stroke of the non-threaded part of the jacking screw, we are left with a possible thread length of only .377 without lengthening the part and the jacking screw. That said, if molded-in threads are financially optimum, I will take the trade-offs and simply lengthen the part to allow for a longer threaded section.

Thanks,

Dan

 

[Pud]

Dan,

30 cents? (maybe China with dubious quality) No chance!

You would require a high performance material, certainly inserted (for wear), blah blah blah....

Moulded thread tools are around £1000 per spindle ($1300?)

I would go for a (simplified) turned part in Al or MS. (And this from a plastics business!)

Harry

http://www.tynevalleyplastics.co.uk

 

[patprimmer]

You can of course make a prototype from bar stock and test and I would strongly advise that course.

I just read 100# load in both directions.

That requires either ultrasonic welded or moulded in insert or welded in retainer for a std metal nut. The cost limitation will be difficult no matter where you make it.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

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

for site rules

 

[dapeake]

I agree. The price, at .30, is a challenge. Especially with my preference to avoid China.

I do believe that a four family MUD type insert mold will meet my needs provided (i) I have the end user assemble the thing (give them three parts, a nut, and the two plastic halves) (ii) fully automatic molding. (iv) fast cycle times.

I presently have several parts made here in the US which are larger and more complex for around .60 to .75, using a single cavity tool. With a four cavity tool, i.e. $1.20 per shot - minus the cost of the 4ea, 5/16-18 nuts, .30 seems achievable.

Prototype is coming soon - I wanted to get this community's input before I go into the tool room.

Ultra-sonic welders: How fast do those cycle?

Thanks,

Dan

 

[CoryPad]

Post-molding insertion is preferred over molded-in thread inserts for adhesion and strength. A typical ultrasonic thread insertion machine can run at 80 cycles per minute, although this is application dependent.

 

[MikeHalloran]

I'd take a good look at molding a rectangular slot perpendicular to the axis, and supplying a brass square nut for radial insertion by the user. The slot could get some snap fingers to retain the nut while allowing it to self-align on the screw.





Mike Halloran
Pembroke Pines, FL, USA

 

[plasgears]

Depending on volume, I would cast the thds in the mold and unscrew the core. This minimizes machining and enhances thd strength. I favor acme thds.

Don't source in China, if that is what you are thinking.

 

[dapeake]

Mike; I like that idea. I had considered a similar idea using a regular hex head nut, but the possibility of rotation of the nut under load turned me against it - the square nut is superior. Any ideas on how to shut off the transverse rectangular core against the head of the axial .312 core on the A (hot) side of the mold? I am concerned about a wiping shut-off there.

Corypad: I am not understanding how the strength of an insert, installed via ultrasonics, is going to out-perform a molded-in insert in terms of pull strength. Would you mind going into further detail on that?

Plasgears: Are you talking about hand-unscrewing the core? If so, I'd be interested in the efficiencies in terms of 2nd op expenses. With a core unscrew - an operator must attend the press to load and remove cores.

Thanks,

Dan

 

[Pud]

There are several ways to automatically unscrew cores: hydraulic motor (long threads)/rack and pinion (short threads)etc.

Just costs more for tooling so benefits depend on volumes required off the tooling vs labour costs for alternatives.

h

http://www.tynevalleyplastics.co.uk