Odd Screw size???

Hi Huys,

I have a Katsu jigsaw. The blades are held in with 2 socket head screws marked 8.8 and HDW.

The TPI is I think 38. The outer diameter of this screw is 3.7mm. I have just matched the thread with one of mu gauges, metric 60 degree 0.7. How do I translate that into screw spec?

Chris

Edited By Chris TickTock on 02/01/2021 18:10:45

It's going to be metric, off the top of my head M4 x .7 pitch???

Tony

I would say that too. 8.8 is a metric thread strength designation and 38/25.4 = 0.67 so I'd start by trying a M4 tap in the threaded hole.

You beat me to it Tony...well done. I was under the impression that a M$ thread would measure 4mm across its thread, obviously not.

Another lesson learned today

Chris

M4 X 0.7 is the standard metric coarse thread, and the 8.8 is medium tensile strength.

Almost certainly metric M4 x 0.7 (4mm nominal o/d, 0.7 pitch is about 36 tpi)

8.8 is a metric high-tensile steel specification, nothing special.

I got so fed up making mistakes counting thread turns I bought a gauge. Much easier!

Dave

Chris -

It's that apparent size reduction that seems to have thrown you.

Thread crests are normally slightly below nominal size. It avoids sharp crests and clears the female thread root, which itself is radiused to minimise stress-raising. The actual reductions are as carefully standardised as the pitches, angles, etc..

An M4 screw I happened to measure only this afternoon, to verify it was indeed M4, was 3.7mm dia., too.

It will be worth you obtaining sets of thread tables to go with the thread-gauges, if you've not already done so.

A quick test of a thread is to assess its fit on a known opposite-number. They should act smoothly together with little shake and no initial or cyclic binding - but be aware that there are a few near matches between individuals in different systems despite their differing thread angles. There is also a fine-pitch Metric range, though these don't seem to crop up very often outside of instrument-making and automotive engineering.

Thanks Guys,

Lessons learned.

Chris

Elsewhere on this Forum, it has been posted that threads are truncated to prevent root/crest interference.

This other post, stated that the truncation for a Metric thread would be 10% of the pitch. So for a 4 x 0.7 the OD would be 4 - 0.14 = 3.86 mm, so your 3.7 is a little undersize. But the thread may be rolled, and so a little undersize anyway.

This will reduce the % engagement, but is unlikely to cause problems. ( If thread strength is important, as in say, yield tightening applications, mass produced rolled threads would probably not be used ).

If it works satisfactorily, why worry? If it ain't broke, don't fix it!

Howard

Howard I have just measured 3 unused metric threads (2 fine, 1 coarse) they all work out in terms of measured diameter to be almost exactly 92.5% of the nominal diameter size.

Not saying you are wrong but to date I would question its usefulness. Needless to say different spec on manufacturing will give some inevitable tolerances.

Chris

Edited By Chris TickTock on 03/01/2021 12:13:53

ISO Metric Threads

In Machinery's Handbook. I was reading a couple of days ago that the crest of the male thread is rounded to an amount of H/8. H being the pitch.

I was making M10x1.00 male threads (in brass). I calculated, using H=1.00, that the outside diameter should be 10.00mm - 1.00/8 - 1.00/8. or 10.00mm - 0.125mm-0.125mm. Answer is 9.75mm. It worked!

Not having the Handbook with me right now, I cannot check, but M4 x 0.75 should yield 4.00 - 0.75/8 - 0.75/8, or 4.00 - 0.09375 - 0.09375. Answer 3.8125mm outside diameter.

Happy to admit that my memory is not so good, but the new diameter is 91.325% of 4.00mm, which seems to agree with previous posts.

Chris,

I merely quoted what had been said a few days ago, on another thread on the Forum. However, your apparent disbelief prompted me to check.

Kempes Engineers Yearsbook, 1982, quotes the proportions for Metric threads as:

H (Depth of thread ) as 0.86603P where P = Pitch.

Truncation is quoted as H/8

If you use your calculator, you will find that 0.86603 / 8 = 0.1082537.

This, as long as you ignore the 4th and subsequent decimal places, matches the 0.108P quoted by Tubal Cain in his "Model Engineer's Handbook"

So the 10% quoted previously is probably close enough for most purposes.

If you can find any BS or ISO standard which changes these proportions, please feel free to tell us which they are.

Howard

Isometric Thread Form

John

Alan typed whilst I was checking..Machinery's Handbook appears to quote the same proportions as Kempe's.

So, a standard 4 x 0.7 thread would be truncated by 0.075 mm, (0.7 x 0.108 ) reducing the diameter by 0.1512, to give a dimension over the crests of 3.8488 mm, or 96.22%.

So, your 3.7 mm threads are undersize compared to the generally accepted standard

Howard

Edited By Howard Lewis on 03/01/2021 13:38:52

Done a bit more digging around. Seems there are grades of fit for commercial uses of threads. Then there are sloppy manufacturers. So I revise my previous post which though empirically correct as the 3 screws measured rendering a 7.5% deviation this probably just indicates the sloppy end of manufacturing tolerance available. I guess for most hobby uses fit is not critical but having said that I will be watching out a bit more.

Also if I have say a metric hole I doubt I could easily tell the fit grade that was made to to match an equivalent bolt / screw anyway. If I tapped a new thread what fit grade would that be? Life could get complicated.

What started out to be a simple question of what thread is this in relation to its outer diameter has grown legs. I am happy to say the 3.7 measured is a sloppy fit for a M4 0.7 thread.....poor fit that it obviously is.

Rgards

Chris

My Machinery's Handbook tabulates BS upper and lower tolerance sizes for "medium fit" metric threads. For M4 on the major diameter these are 3.978mm and 3.838mm resp. ie 0.14mm spread.

Effective diameter is also tabulated, having a closer spread of tolerance of 0.09mm. but on minor diameter the spread is again 0.14mm.

Another table gives larger, "free fit" tolerances but sizes less than M5 aren't quoted. Perhaps "free fit" becomes "drop through" in the smaller sizes.

The fit that WE can make is determined by

The Tap that we use, after drilling the hole (How accurate a hole does the tapping drill produce? Size and symmetry or otherwise of the lips )

The Die that we use, and how accurately we set it.

I always adjust my Dies to a commercial thread.(preferably a good quality bolt ).

If you read the comments by folk such as G H Thomas, Tubal Cain etc, it will be clear than for MOST attachment purposes thread engagement has a limited effect on the ability to clamp parts together, or on thread strength.

You only need to start using closer classes of fit if you are seeking greater precision (Micrometer barrels and screws come to mind ) or seeking a closer relationship between applied torque and applied load, or when yield tightening.

The thread fails (strips ) when the load is so great that it shears the thread from the core material, until then it applies a force to the fixing which causes elastic elongation, (and compression of the parts being clamped by the fixing ).

On aero engines, the clamp load on Big End bolts was determined by measuring the extension of the bolt. But the characteristics and material of the bolt were very carefully controlled!

Yield tightening takes this a step further, to maximise the load obtainable from a fastener by taking it just into yield so that it remains very slightly elongated, after the load is removed. A "W" range 1/2 UNF bolt will provide a clamp load of 9 tons when just into yield, but will be about 0.002" longer afterwards.

Taken to excess, the fastener yields and just before tensile failure, starts to "neck" But this should have become obvious as further e rotation produces little or no more clamp load or torque requirement.

If there is enough frictional force between the threads, it is possible that the fastener will "shear off" and fail in torsion. It is not really a failure in shear, since the load is angular rather than transverse, as in true shear.

Combining loads will produce failure earlier than if just one load is present. Cyclic loading takes it into fatigue, so that a fastener or any component which withstands a constant applied load could eventually fail if the load is cycled on and off. Hence "fatigue limit" often taken as a minimum of 10>7 cycles without failure.

Fairly slack fitting commercial nuts and threads can apply a considerable load before one or other part of the threads fail.. A thread is only an inclined plane wrapped round into a circle, to provide a mechanical advantage.

Think of a screw jack for changing wheels on a car. The handle rotation uses a great linear movement of the effort to produce a small movement of the load.

Howard

These are the relevant tables in the 1981 standard for metric threads.

Martin C