Pull Out Force for Metal Inserts

Dunno...but experience has shown me

Those cold press in ones are useless, unless you like squeaky screws which bind and pull out the insert.

...use heat or ultrasonic or molded in ones...

The pull-out force depends on the type of insert and if it was pressed in or in the mold at injection. If you give those informations it is possible to gieve you a relationship. It depends as well of 2 other parameters : type of thermoplastic material and geometry of the place where the insert is introduced. Those data are also determinant. All formula are only giving an indication since in the case of pressed inserts the bore is manufactured with tolerances and the force will depend on tolerances as well. I hope this will help you to have as i presume a quantitative indication.

There is no formula for this as there are too many variables in material and inserts and methods of placing the inserts in to the material.

I would suggest that you set up a test fixture, (it does not have to be very complicated,) and test the pull out force of the different inserts you are thinking of using.

I manufactured some polycarbonate parts that had up to 24 inserts per unit and due to the lack of information on pull out and torque resistance I set up my own test.

In less than a day, I had my answers and over the years it has been a trouble free choice.

Nice one...sensible engineering, a breath of fresh air!

I sometimes think we lose sight of this sort of approach...simple quick and accurate.

Del

I partly agree with you. Of course there are many factors but some are major and other not. In engineering ther is a method called "cut and try" which gives good results takes a lot of time and does not permit the usage of experience for an other similar case. Ther is a method which is based only on theoretical analysis which i personally consider stupid and dangerous. And there is a third way which based on a simple physically correct model tries to predict values and allows after tests to correct the start assumptions and use the experience for other similar problems. This is on my opinion and based on a quite long experience in development of new products an optimal approach since it gives the possibility to reduce redundancy in tests. The problem with tests is that especially when many parameters are involved the dispersion of results can be quite important. If the number of samples is limited and by chance (or the contrary) most deviations move result in a direction the obtained value can be too optimistic and generate troubles when in production some deviations go the other way. Tests should be done keeping this in mind and as far as possible the samples must be so that extreme situations are generated in all important directions.

That is very true, I fully agree.

...but in the time it took me to read that, I could have built my rig and done the tests! (just teasing)

Like most thing is it using appropriate tools, tests etc.

I think it's called experience!

There was a thread a few days back ..some chap asking how to work out if 4 hinges could hold up a 200lb door in cosmoworks!!! Completely innapropriate as the screws and what it is fitted to was far more pertinent.

Cheers

Del

I also agree that what you call experience and what i call "feeling" and sound human thinking are extremely powerful engineering tools unfortunately used less and less since simulations are not any more considered as virtual but as REALITY! I met young engineers who believed more in computer results than in a logical plausibility analysis.

In fact they did a units error before they put the values in the soft and the result was totally wrong but if the computer said it it should be so. It took me a very big effort to explain where the error was.

My favourite was a young mech' eng' I employed ..he had to check a computer controlled piston (breath simulator).

He came up to me like the sky had fallen in saying the timings were out!

I went over to the machine, and started it on an 8 second stroke..

I counted 1 and, 2 and, 3 and.....8 and it stopped.

'That's fine, nothing wrong with that' I said walking off.

About half an hour later he came back tail between his legs....

He'd borrowed a stopwatch from prod' eng' which worked in decimals of an hour for time and motion studies!!!

I'd never heard of such a thing!

But it just goes to show you need to actually understand the magintude of your units. It took a week for his colleagues to stop teasing him!

Del:

This reminds me of an old saying (don't know who it's attributed to) "Before you make a measurement, make sure your yardstick is three feet long."

My Dad used to say, measure twice, cut once.

Hank

My Dad used to say, measure twice, cut once.

Good advice for a carpentry apprentice to hear.

I agree. We do a lot with honeycomb composite panels and there's so many variabes (skin, honeycomb material, thickness, insert installation error etc etc etc etc) that testing is really all you can rely on and then just give yourself a reasonable margin for error.