is my thing big enough (for general screwing)

May 17, 2009 6:53 PM

will a M6 cap screw allow me to use a higher torque spec than and M5 cap screw

in an application, it is important to get a good fastened seal between 3 substrates.
an M8 bolt it done up to secure two substrates.
then a third substrate is attached with two M5 cap screws that pass through all 3 parts (middle substrate has through holes).
the problem is, when the cap screws are done up, the M8 can loosen.
solution 1 is to use a spring loaded M8.
but another solution that has been discussed is to upspec the M5 fasteners, thus making the initial joint less critical.
will this allow the inc rease in torque spec that i am looking for?
i dont have access to a machineries handbook so not sure where to look.
(sorry for such a boring question)

posted by edtut to science & nature (10 comments total)

It's not a boring question. I just had to think about what all those words meant.

First of all, from here, the torque specs for M5 and M6 socket-head cap screws:

M5 x .80
Grade 8.8....................................4.13 lbs./foot
Grade 12.9....................................6.9 lbs./foot

M6 x 1.0
Grade 8.8.....................................7.1 lbs./foot
Grade 10.9.....................................11 lbs./foot
Grade 12.9...................................11.6 lbs./foot

So: I don't know what increase in torque spec you're looking for--I don't know how to evaluate the torque rating of spring-loaded cap screws, unfortunately; you probably know more about this than I. However, I *can* tell you that, assuming you're on 12.9-grade M5s, you'll gain 4.7 lbs./foot by switching to the same grade of M6.
posted by koeselitz at 7:58 PM on May 17 [1 favorite]

I'd definitely change the M5 to a bigger screw, and then, I'd lose the M8 entirely. I'm assuming that you can't get access to the M8 when the third part is attached.

The problem -- you preload and torque down the M8 correctly. Then you compress the substrate it's bearing on by bolting on another piece, which removes the preload and the M8 spins.

Just going to M6 and torquing down more is going to make that problem worse. So, get it so that the two screws attaching the last substrate can provide enough torque to clamp the entire stack -- then lose the middle bolt entirely.

If you can torque down all the screws to proper preload, the problem won't occur. But, just like when you attach a wheel to the car, you need to be able to load each fastener until the parts are settled and compressed, and each fastener has the proper torque.

So, that buried M8 is always going to be a problem. Get access to the head, so you can retorque after the M5s go in, or get rid of it and use whatever cap screws you need to clamp the entire stack to tolerance.
posted by eriko at 9:25 PM on May 17

eriko you didnt consider and i didnt mention that the first attachment is an assembly aid.
so not sure it can go completely, but i agree.
by my calcs you can probably increase the securing torque by about 75% by moving from M5 to M6.
if it is left to re-torqueing, i think it will get missed on the line eventually and the problem will resurface.
posted by edtut at 9:56 PM on May 17

Could you, like eriko suggests, englage the M5 bolts so the M8 isn't required, then because the M8 is an assembly aid only (rather than being required to hold the assembly together) you could secure it with some loctite.
posted by Mike1024 at 2:10 AM on May 18

Can you replace the M8 with a pin of some sort, providing the assembly aid you need while not having a role in the structure of the final assembly?
posted by dg at 3:44 AM on May 18

Unless this is something that requires regular assembly and disassembly, why not put a drop of blue loctite on the M8 cap screw and call it good?
posted by Kid Charlemagne at 5:32 AM on May 18

Like edtut said. The first bolt is always going to lose preload when you tighten the second bolts down.

The load is proportional to the area of the bolt. (Roughly speaking the force applied by a bolt is 5*Torque/diameter (Nm & mm) for a black bolt or 6.6*T/d for a plated bolt.) So, if the same grade bolts are used, the M8 is applying 64 units of preload and the M5's apply 2*5^2=50 units. The preload on the M8 is seriously reduced. A spring washer might be enough to stop it coming undone, although IME loctite is better.

If absolute load isnt a problem then you could use a higher grade M8. It would lose proportionally less preload and so would not be so likely to come undone.

Having said all of this, if your bolt is coming undone then preload may not be the problem. If your assembly is vibrating or heat cycling then the I'd consider using proper lock washers (I've good experiences with Nordloks) or the appropriate loctite. Your longer bolts will be less susceptible to this than the shorter one.
posted by BadMiker at 7:50 AM on May 18

oops.

Like eriko said.
Sorry
posted by BadMiker at 7:51 AM on May 18

thanks guys. actually the first joint is also required to be leak tested, then delivered to another line, so the M8 cannot really be removed.
the nordlock looks interesting, also considering a belleville washer.
but badmiker, the problem is actually two fold - preload is definitely an issue, as we have parts come loose just at assy (probably not torqued to spec, but still representative of servcing). but also, the parts are in an engine bay, so over time the joints are subjected to vibration as well.
not sure if the nordlock will solve the preload problem.
posted by edtut at 5:06 PM on May 18

Then you have 3 problems!

Nordlock or loctite would help with the loosening due to vibration
A higher grade M8 will be less likely to come undone when you tighten the M5s
But you still have the quality control issue!

It sounds like you need to make the assembly much less susceptible to quality variation.

The relationship between torque and force in a bolt is very dependant on the friction of the threads. Greased threads, zinc plated bolts, loctite, grit and rust all have a significant effect on the resulting force from the same applied torque. Is it possible to tighten up the assembly method so it's much more consistent? Use plated bolts into a clean thread?

Making the bolts longer would mean they are much more tolerant of changes in preload. Likewise, more bolts of a lower grade would give more redundancy. Rotabolt is a system we have considered as a way of monitoring the torque. It would add expense to the assembly but would give an inherent self-check to the torque.

This is very good.
posted by BadMiker at 7:42 AM on May 20