Press Fit

[Peter Simpson 3Participant @petersimpson3]

Hi all,

I’m in the process of machining some parts out of EN8 mild steel

Basically it’s a tube with ID of 0.845″ and a wall thickness of 3.5mm. I want to make a plug to fit in one end of the tube. The plug needs to be a press fit, I do not want the plug to be removable under normal conditions. I have a hydraulic press to ensure the plug can be pressed home.

The question is what dimension should the OD of the plug be ?

[old martParticipant @oldmart]

You also need to state the length of the plug and bear in mind that the OD of the tube will increase where the plug sits.

[SillyOldDufferModerator @sillyoldduffer]

Tubal Cain says for holes between 1/8″ and 2″:

• The hole should be dead on and the shaft adjusted up or down.
• A force fit is +0.75 thou per inch of diameter, which makes a 0.845″ + 0.00634 = 0.85134″ plus a constant
• The constant is 0.0005, so I reckon you want a 0.852″ plug

If it needs to be very secure consider a shrink fit, where the plug is made rather bigger, and the size of the hole increased with a blowlamp.

Dave

[JasonBModerator @jasonb]

I think Dave will be banging that one in all day

Try 0.846 plug into an 0.845 hole.

[bernard towersParticipant @bernardtowers37738]

A 7 thou interferance fit I think not

[garyParticipant @gary44937]

make it a push fit with locktight

[Paul KempParticipant @paulkemp46892]

Oh I laughed, well done Dave!  Others beat me to it!  You will be lucky to get a plug some 7 thou bigger than the hole in, would need to be over 8” in diameter to have half a chance.  Look at your decimal points, TC is saying 3/4 thou per inch.

I agree with Jason’s size, that won’t be falling out in a hurry.

Paul.

[Tony Pratt 1Participant @tonypratt1]

About .847″/.848″ plug into a .845 hole will give a tight fit, slight complication being the tube is likely to swell away from the plug. The .007″ interference quoted above is I believe incorrect?

Tony

[Peter Simpson 3Participant @petersimpson3]

The plug will be about 0.750″ long, the OD of the tube will be machined to final dimensions, after the plug is installed.

[JasonBModerator @jasonb]

a thou will be plenty, ease the leading edge of the plug so it does not gouge the hole

Not only did Dave get his maths wrong, He does not have the experience to know the answer was way off.

Check your ID measurement as digi callipers can under read a bit in a hole and tube is not always round

[Peter Simpson 3Participant @petersimpson3]

Thanks guys will go with 0.001″

[Howard LewisParticipant @howardlewis46836]

Fully agree with 0.001″, and that would be a tight fit, liable to distort the tube slightly.

When I fitted the mild steel column into the 6″ dia, 1″ thick C I base of my comparator, I used a 0.001″ interference in a 1″ reamed hole, and used Heat and Freeze technique. The column slid into the reamed hole, but the whole thing can now be picked up by the column.

Howard.

[Clive FosterParticipant @clivefoster55965]

For press fits I prefer the self gauging taper method to the traditional parallel system. Most especially if things have to be assembled cold.

Idea is to make both male and female parts with a shallow taper whose half angle (taper on one side) gives the desired interference fit over the full length of the taper part. If both male and female tapers are dead nuts accurate the male will enter the female freely up to half the press fit depth. Pushing fully home gives the desired press fit all along the part.

The taper makes assembly much easier because the parts are aligned by the taper before you start pushing. The actual fit increases steadily from zero at the initial half in point to full when completely home so the effort is much less than with a conventional parallel fit which is full fit all the way down. Much less chance of damaging the parts and, in this case where the female is tube much less risk of expanding it during the process.

Setting up for the shallow taper can be a bit tricky. Best way is to use the a decent travel on your taper turning attachment, or topside if you’ve not got an attachment, and measure the actual deviation between the ends in thous per inch or its metric equivalent. A dial gauge gives accurate results quite easily but creative use of the cross slide dial is pretty good.

Obviously machine one taper on the back side of the work and the other on front to produce two matching tapers automatically. Once you get close enough to size for things to start entering shift the cutting tool longitudinally  using topside parallel to the bed if working with a taper turning attachment or saddle against a bed mounted dial gauge if using a set over top slide. Due to the shallow angle of the taper the longitudinal movement needed to take out the last few tenths is large enough to be easily measurable with the sort of equipment we have in the home shop.

Most folk don’t have the gear to accurately measure and make a parallel shaft and bore a true 0.001″ different. Getting a press fit of some description merely takes care but how close the result really is to 0.001″ press fit is going to be anybodies guess.

I have got the machines and gear to do it but really wouldn’t like to take bets on coming up right first time from a cold start with only one pair of parts to play with from a standstill. Easily between 0.0005″ and 0.0015″ for sure, if I don’t have a brain fart mid job, but exactly what Inspector Meticulous would find if he separated the parts I don’t know.

With the self gauging taper method it’s pretty easy to get the longitudinal shift within a couple of thou of half way so the error in fit should be infinitesimal.

The major disadvantage of the self gauging method is that it’s a new skill to practice and dial in. Parallel press fits are  “just” a standard turning and boring job done really, really carefully.

That said I’ve gotten entirely satisfactory results using parallel shafts and bores before I found out about the self gauging method. But I always used heat to expand the bore so the shaft just dropped in. Clearly the resulting fit was tight enough.

Clive

[bernard towersParticipant @bernardtowers37738]

Why not machine a gauge pin of 0.845 then bore tube to fit pin, then make insert plus whatever and bobs your!

[Alan CharlestonParticipant @alancharleston78882]

Hi Peter,

I’m always a bit wary about press fits. You need to be crack on with both measuring the hole and machining the plug. What I tend to do is to make the plug so it only just goes into the hole, then put 4 centrepunches about half way down the part of the plug going into the hole at about 90 degrees to each other. Easy enough to do so might be worth trying.

Regards,

Alan C.

[bernard towers]

On bernard towers Said:

Why not machine a gauge pin of 0.845 then bore tube to fit pin, then make insert plus whatever and bobs your!

I would concur with Bernard on this one, but if you make the other end of the gauge 0.844 and one end goes into the bore and the other end doesn’t you know within a thou how big your bore is. You can then make your plug 0.846

I would always make the first 1/8 of the plug to 0.844 to ensure it enters the bore square.

Andy

[larry phelan 1Participant @larryphelan1]

Dont mind them Dave !   I,m with you, hang the experts !

If it does,nt go in, beat it in !

[BazyleParticipant @bazyle]

Interesting info about the taper method. Of course the surest way is to aim for a sliding fit and leave it for a week only to find the stupid thing has rusted into place and nothing you can do will free the parts. 🙂

[Alan Charleston]

On Alan Charleston Said:

Hi Peter,

I’m always a bit wary about press fits. You need to be crack on with both measuring the hole and machining the plug. What I tend to do is to make the plug so it only just goes into the hole, then put 4 centrepunches about half way down the part of the plug going into the hole at about 90 degrees to each other. Easy enough to do so might be worth trying.

Regards,

Alan C.

Exactly.

Probably easiest way for “ordinary guy or gal” would be to ream the hole with a fixed reamer and make the pin to an appropriate external measurement. Assuming the reamer is made on size and your micrometer is accurate you are down to only one potential error. On an easy, routine, machining and measurement job too needing only that bit of extra care to get really right.

But who is going to buy a reamer for one job?

Easy to overlook the importance of accurate measuring gear. Naively a digital vernier with inside and outside jaws ought to handle the measurements. In practice the inherent accuracy limitations mean you are snookered before you start. Its not exactly easy to get a bore measurement exactly across the diameter to with  0.001″ either. Typically digital calipers claim ± 0.001″, 0.04mm. If it’s Mitutoyo I’ll believe it. LiDL special or affordable brands er “Ahem”. Actually my disposable LiDL special is pretty good but not up to this sort of job, eats batteries too.

I’ve been collecting crap for a long, long time so I have tenths reading micrometers, Jo blocks to verify calibration (new sets from Weber) and a half decent hand with telescopic bore gauges. But I’d still not care to guarantee to routinely being able to beat a ± 0.0005″ error band on the 0.001″interference fit. It will be better but how much?

There are darn good reasons why ± 0.001″ is reckoned to be about as tight as can routinely be asked off a lathe or mill. ± 0.005″ is a darn sight more achievable for routine work. In the factory high accuracy comes from the grinding shop.

Out in our world make it to fit rather than make it to numbers has always been the way. Although CAD and CNC is slowly changing that. Plenty of people have produced functional interference fits without sophisticated measurement equipment using care and good workmanship. But there is a difference between functional and made to tolerance.

Which in our world can be largely irrelevant. If it works it’s good. But it might take a couple or three goes.

The nice thing about the self gauging taper method is that it regularises making to fit with close tolerance using pre-setting dimensions of more amenable size.

Clive

[JasonB]

On JasonB Said:

I think Dave will be banging that one in all day

Try 0.846 plug into an 0.845 hole.

Whoops, it’s a mistake, plus I forgot to add the usual warning about my dodgy maths!

In mitigation:

• I said where the calculation came from, so it can be looked up.
• I included the working, which shows where the error slipped in (a missing 0 moved the decimal point)
• Explaining the method to a beginner is far more important than getting the sums right!
• Supports my argument in another thread that it’s always necessary sanity check other people’s plans.  Especially mine!

I made a mistake, am pleased that it’s been corrected, and am not going off in an aggrieved hump burbling nonsense about my grey hair being disrespected.  Shouldn’t be seen as an opportunity for childish yah boo point scoring.  Fortunately, this being a polite forum, we never have any bother with that…

Got a much more difficult problem to solve today; a bricked laptop!  And why a 9Gb backup won’t fit into a 32G memory stick.

🙂

Dave

[Graham MeekParticipant @grahammeek88282]

According to BS 1916C August 1954. (Taken from my first year apprentice Zeus Book)

A Press Fit in Ferrous Parts for holes between 0.71-1.19″ is Hole (H7) the limit is 0 to + 0.0004″ and the shaft (p6) the limit is +0.0008 to + 0.0013″. Take the measured hole as 0 or Zero and work from there.

Anything beyond this requires heating the hole, and a whole different set of limits come into play then.

Regards

Gray,

 

[Clive Brown 1Participant @clivebrown1]

In this thread there have been mentions of fits centering on the recommendation of around 1 thou per inch diameter interference. This is fine, but in this particular case I think that it’s worth taking into account the fact that the “hole” is in fact a tube which will expand much more readily than a solid block of steel. A slightly heavier interference is therefore practicable.

Incidentally, unless this assembly is for elevated temperatures, I’d consider Loctite.

[bernard towersParticipant @bernardtowers37738]

If you drill a hole to press fit something in won’t that be a tube?

[MacolmParticipant @macolm]

I usually cheat with press fits. In this sort of situation I would machine a pilot diameter maybe 1.5mm along the plug until the tube is a close push fit. Now it is easy to machine the next bit, bigger by the desired interference for the press fit. In this case, one thou interference may well be quite suitable because the tube is relatively thin wall, and thus expandable. Depending on the desired length of the plug, the pilot can be machined off once it has served its purpose, or retained if it will do no harm.

Note that if you are press fitting a substantial pulley to a solid shaft, it is a good idea to stick to the proper BS 1916 sizing, unless you have a gigantic press, and a supply of good luck!

[Neil LickfoldParticipant @neillickfold44316]

So you have a MS tube with about a 21.46mm inner bore and want it plugged. Most of the press fit will just expand the tube section. Anywhere from Ø0.03mm to Ø0.05mm will work on a bored hole. A drilled hole will need more if the hole is not round. Using a smear of castor oil on the inner of the tube and the plug will aide in its assembly. If the tube is heated to 180c, castor oil starts to smoke, then quickly put the plug in while it is warm.

As mentioned, you could make it a neat slide fit, and then use Loctite 648  or if it gets warm, under 230c, then use Loctite 620 high temp. Either system will work.

[Author]

Posts