Phillips head torque effects 3

[Vayenne]

Hello,

I am trying to write a new assembly torque specification for our product, but I'm running into a quandry that Google is not helping me solve. This is not a high-stress structure, but it is a primarily sheet-metal product that needs to do a sufficient job of holding itself together through shipping and daily use. We've had problems with fasteners loosening and falling out due to insufficient installation torques, so we're writing our assembly facility a spec.

The problem is that we are currently (I'm trying to change this) specifying a lot of phillips head button screws. I can find plenty of specs for torquing cap screws and hex screws of these sizes, but I think that a cheap, standard phillips head screw must change the failure mode. It seems to me that the phillips contour will strip out long before a standard hex head preload torque is reached.

a) Is that a correct presumption?

b) If it is, then does anyone have a guide or an idea of what the reduction is to transition from a standard torque spec to a phillips head torque spec?

These are all steel-on-steel fastener interfaces.

Thank you,
~Vay

 

[KENAT]

Vayenne, you say steel on steel but what about the depth of thread?

Are you talking screws into nuts (or 1-1.5D threaded holes) or sheet metal screws?

We've had similar problems but the fact that we have screws into so many different 'threads' (sheet metal, nuts, studs, aluminium, stainless ...) as well as a range of heads means that I'm very concerned about using standard 'rule of thumb'.

The manufacturing engineering manager came over all excited the other day because he'd got a sheet from a vendor with the torques they use, but it didnt' take account of all these variables as far as I could see.

I'll be interested to learn more too.

KENAT, probably the least qualified checker you'll ever meet...

 

[Vayenne]

Kenat,

Ah, good point. We've got a mix, mostly screws into PEM nuts and regular nuts, but some applications involve screws into sheet metal and other materials (plastics, PCB mounts). We're planning a spec value for screws into PEM nuts and regular nuts and a lower spec for screws into plastic and sheet metal.

Thus far, our in-house measurements are starting to suggest that there doesn't need to be a distinction between Phillips head installation torque and other hex heads. The Phillips heads just have a tendency to fail after a few installation / removal cycles, far sooner than hexes of any sort.

Thanks,
~Vay

 

[KENAT]

Yeah, the different head types wasn't something I thought a lot about, I was mainly thinking about the thread stresses etc.

I do know there are some here who don't like small button heads (hex) as they think the hex gets rounded out too often and it can cause problems.



KENAT, probably the least qualified checker you'll ever meet...

 

[CoryPad]

It is not common to use a cross recess for a fastener that is loaded to a large fraction of its pretension limit. Perhaps your application is not that demanding, but your testing is proving the poor performance of the cross recess drive feature. If you are able to fail the head after applying a tightening torque a few times, then you are deforming the drive feature. The deformation is "absorbing" some of the applied torque so that less is going into pretension. You won't find much public data regarding torque resistance of cross recess fastenes. I recommend you switch to a hexalobular design (also know as Torx) - it has much higher torque resistance, and you can get actual engineering data for them. One source is Acument:

http://www.acument.com

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[MintJulep]

a) Is that a correct presumption?

No.

There is no such thing as a "standard torque" (although too many people believe that there is). Mainly because torque is not the desired goal with a threaded fastener. A known tension (preload) is the goal. There is a tenuous relationship between torque and preload, so measurement of torque is often used to estimate the achieved preload in an assembly environment.

The nature of the torque - tension relationship for any particular application is not in any way a function of the drive feature of the fastener.

However, if the drive feature fails before the desired preload is achieved, then the joint will fail in service.

Bottom line: If you need your screws tighter, don't use phillips.

 

[MikeHalloran]

The Philips cross- recess has two big advantages:

- In production, it can be used with auto- feed drivers because it's forgiving of slight misalignment.

- In a less than pristine application environment, e.g. on a dirtbike, it speeds service because you don't have to clean every last little bit of dirt out of the recess in order to engage the driver for disassembly and reassembly.


Ultimate drive torque is not one of its advantages.







Mike Halloran
Pembroke Pines, FL, USA

 

[TenPenny]

I seem to recall a comment that I read in, I think, the history of the Roberston screw head, that one advantage that the car manufacturers liked about the Philips head was that it 'cammed out' when the torque increased.

If you need your assembly to use higher torque, I think your first step should be to go to a Robertson or Torx head.

 

[CoryPad]

When it was first developed, the Phillips cross recess was designed to cam out when it reach a certain point so that it wouldn't overtighten wooden joints. This is now a huge disadvantage for modern joints that need high pretension and must not have surface damage due to mobile assembly tools.

The idea that a cross recess is better than other drive styles with respect to off-axis driving is completely untrue. A nice comparison was made recently by Archetype Joint and shows the maximum angle for a cross recess is 2[°], while a Torx can be driven at 14[°], and a hex can be started at 17[°] but finishes at 11[°]. The study is available for free here:

http://www.archetypejoint.com/comparisontest/comparison_test.cfm

You have to sign for a free membership to view the document.

Regards,

Cory

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Twoballcane]

If I can make a suggestion, if you get a torque wrench with an analog dial and tighten a few screws. You should be able to come up with an average of torques values that “feel” right for the application.

If there are no requirements that you have to use Phillips head screws, why not got to socket caps? Socket caps are my first choice. You can check out Mcmaster.com for a variety of socket caps.


Tobalcane
"If you avoid failure, you also avoid success."

 

[Vayenne]

Thank you all for your help and advice. I now have the spec values that I wanted, and excellent points to consider as well.

As a little aside to this discussion, many of you have pointed out the weaknesses of Phillips head fasteners when it comes to re-usability and installation control. I couldn't agree more, and I have already gotten approval to replace our Phillips fasteners with socket head and hex head versions. I will be working with my fellow designers and production facilities to phase out cross-head fasteners in favor of hex or torx-based interfaces.

Thanks for the assistance,
~Vay

 

[KENAT]

Well, seems most of you have confirmed my concerns that the application is what's important and that if you have a lot of slightly different applications then using a 'general rule of thumb' torque is probably a bad thing.

A guy at my last place used to jokingly specify '2 white nuckles' now I just need to know that in standard units.;-)

Vayene, just a thought I had for your application, can you use threadlocker (or maybe patched screws) at all? If you're using the pre-load to mainly stop it from un doing rather than any other reason would this work?

KENAT, probably the least qualified checker you'll ever meet...

 

[drawoh]

Vayenne,

On my first pass, I do not care what kind of socket my screw has. I calculate screw torques based on 90% of proof stress or yield stress. I have done this for Phillips screws, and had no problems torquing them down. As far as I know, I have had no problems with them torquing out either.

JHG

 

[HDS]

Another advantage of cap screws is that they are often easier to obtain in known grades such as SAE grade 5. With plain screws it is harder to identify the quality of the fastener.