Replacing conventional ordinary washers with hardened washers ends bolt failures

Hi SB

I still don't see how hardened washers will obviate a failure on a bolt. Of course the torque setting will be more accurate as there won't be any galling to increase the friction but why else should it make a difference?

LJ, What type of joint was it? What type of failure?

Chas

As per my experience with the fasteners - when we torque the bolts (or nuts) - the bolt head bites into the washer (and the component if there are no washer) - this can be physically also seen.

Under these circumstances a huge amount of torque is lost int the machining of the washer.

And as the joint compression is lost, the failure of the fastener is the obvious result.

The hardened and ground washers are going to provide a nice sliding and not easily machinable surface so you are more likely to have a better retained pre-stress.

Even when the bolts are stretched (mechanically) the deformation of washer (metal squeezing out) is likely.

In these cases we prefer to have special thicker washers - since the pressure cone on the flange will now have more surface area (remember most of the cases the flanges esp if these are not pipeline but components) are likely to be soft like the unhardened washers. It is always better to transmit the stress through a bit more surface.

Under these circumstances a huge amount of torque is lost int the machining of the washer.

And as the joint compression is lost, the failure of the fastener is the obvious result.

If joint compression is lost, then bolt tension is lost, and failure of a fastener from low tension is anything but obvious.

(I assume this is why Laughing Jaguar made the post -- his expectation would be that the hardened washer would, as most agree, reduce bolt head friction, thereby increasing actual bolt shaft torque and bolt stretch and tension. It is very hard to make the case that 120,000 psi steel is more likely to fail at 60,000 psi than at 120,000.)

An ordinary bolt torque chart, such as the one supplied in one of your links above, shows the effect of friction: dry fasteners require more torque to achieve equivalent clamping force (as compared with lubed fasteners). So in Laughing Jaguar's case, if 3/8 in (grade 5, fine thread) bolts were failing at 35 lb ft dry, then they would fail at 25 lb ft lubed. Following the same logic, if they were failing at 35 lb ft with soft (high friction) washers, they should fail at perhaps 30 lb ft with harder (lower friction) washers. This is the reverse of his observation.

I'd postulate instead, that the soft washers were acting to reduce, not increase friction. For example, if they were becoming conical when torqued, with the bearing area restricted to a small ring around the bolt hole, (rather than to a large area around the bolt hole, as provided by a harder washer) then the torque required to overcome friction would be lower, because the radius out to the friction point from the center of the bolt would be smaller. I'd guess that the mechanic saw that the washers were polished near their inner radius, and reasoned that stiffer washers would better spread the load and increase friction. Thus, a given head torque would be less likely to stretch the bolt to its breaking point (because the given head torque resulted in a lower shaft torque).

If this were the case, then "torqued to 50% higher clamping pressures" in the original post would have to read "torqued to 50% higher values."

Another possibility is that the threads in the part were not tapped to sufficient depth, and that the thicker hardened washer kept the bolt from prematurely binding in threads not fully formed.

I think we are reading the question differently. Because he didn't claim otherwise, I assumed that Laughing Jaguar was talking about an ordinary joint, rather than one using true "prestressed fastener" technique (in which the stress is obtained by measuring the strain with a micrometer, for example). The fact that they were using mild steel washers would support this interpretation, I think.

Also a key, I think, is that he says the failure was upon assembly, not in service. As you say, good joint design would ensure that, in service, the joint friction (as a result of bolt load) holds the joint stationary rather than relying on the bolts in shear (which, given bolt hole clearances would result in 1. a very sloppy joint, and 2. the possibility of shearing off the first bolt that is required to accept all the shear load -- it is very unlikely that all the bolts will accept the shear load at once, if friction has failed.)

So, as the stress is applied in tightening the bolt, the tension on the bolt would go up through 28T on the way to the correct pre-stress, 42T. It would be a defective bolt that would fail at 28t at assembly, i.e, during the stressing process, unless the threads are galled, and the failure is in torsion rather than in tension (but even that requires that someone is not reading the torque wrench).

For a given torque wrench reading, the bolt tension will be higher if bolt head or bolt thread friction is reduced. (In the extreme, weld the bolt head down before torquing, and no reasonable amount of torque will tension the bolt at all -- the bolt would not then fail upon assembly, although the joint would fail, and the bolts would also likely fail in service.)

This is the reason there are the dry and lubed figures in torque tables: torque a lubed fastener to the lubed spec, and you risk overtensioning it, or having it fail in tension at assembly.

So, in service, of course improperly tensioned bolts can fail, because the bolt is used in away unintended. In cyclic loads, the bolts can also fail from tension fatigue: it is far better to tension a bolt once to 100,000 lb than millions of times to 20,000 lb.

But at assembly a bolt that should be tensioned to 42T has to go through 28T on the way, and it will only fail at 28T if it is defective.

It can also be seen from the design of joints, that only a fraction of the variable pressure only (based on the joint stiffness) is handled by the bolts over and above the prestress. ie if the joint load is varying by say 30%, the bolt loads may vary by 10% only.

You are getting at the idea of what has come to be called a prestressed bolt (although any properly tightened bolt is prestressed, I'd say). In connecting rods, the forces are very high, and the load change is frequent, so the idea is to tension the bolt enough that there is no cyclical variation in bolt length (and hence bolt tension).

Chatter.......chattering, and bending wire back an forth until breaking, fatigue failure analysis

"Embarrassed by the simplicity of the solution they tested the cure in a lab and found that the same fasteners, in the same assembly, using the same procedures, could be torqued to 50% higher clamping pressures, without failure, using hardened steel washers."

when using soft washers what was holding/preventing from torquing to same 50% higher clamping pressure?what was failing when u tried to torque to 50% higher clamping pressures

"We put the old washers back and the fasteners failed at half the load!"

What was nature of failure at half the load? do u mean tightening torque by "load"?where was the torque measured at the wrench? or by measuring strain induced in bolt or by measuring clamping pressure?

I'm going to hazard a guess that the harder washer, being stronger and therefore stiffer was distributing the clamping forces better instead of bowing and causing a localized stress riser on the head of the bolt. I've noted that API flanges for instance specify a very large clearance hole that, because the remaining contact area under the bolt head is reduced, generates very localized stresses in both the flange and the bolt head.

Dear L.J.

sing off the shelf washers, is a common mistake,and everyone seems to do it,, even me.

Your answer is of the shelf washers are made of mild steel, AMS /SAE 1020 steel, which yield a low stencil strength, with out enough carbon etc, to heat-treat. Off the shelf guys are used mostly for fabrication so the bout or nut doesn't crush into the materials fasting. these mild steels ranging from 1016 tru 1030 are consider soft steels.e tested using the Bernell hardness test, and the tensile strength if quench in oil with will develop about 96,000 pounds per sq. in. (SAE 1035) and a yield point of 65,000 pounds per sq. in. Yaw are you with me? They are good washer for some and most general use, but not recommend for strength. But like I said we all use them because there easy to get an really cheap to buy.

Now if you want to have HT washers, you can make them or buy them, what ever. Not knowing the mechanical stress loads, one should start with a washer of a a grade 3 with can be bought at a nut & bolt store. They start at a SAE 1035 an go up in grades to about 5 I think. The best idea would be to buy from tool catalog washer the and bolts.

Furthermore, you can make your own, ( which I have) and solved most all my fixture problems. The store bought washers that are use in fixture use are about a SAE 3130 and the tensile strength is 165,000 pounds per sq in and YIELD at 147,000 lbs per sq in. and have a tensile strength of about 35 know as C35.

I used 4140 QT., that had a tensile strength of 180,000lbs. per sq in. and had a hardness of 41 to 42 on the c-scale. Yaw I had allot of 4140 steel laying about.

Let me know what if you have any more questions. /And good Luck METAL MAN

We had bolt failures under high stress conditions due to the washer face not being perpendicular to the bolt. when the bolt was torqued the head face would conform to the washer face causing a rise in stress on the short side. Failure was catastrophic because it was on the front suspension of a +200 ton mining truck. I would suspect your soft 1020 washers were crushing and causing this type of failure. Too bad you don't have one of the failed bolts left.

The orientation of the hardened washer is critical, for your dump truck front struts especially. Your manual will show how to fit the hardened washer, the stamping burr must be away from the bolt head. Other types of hardened washer overcome this problem by chamfering the inside and outside of the washer on the bolt head side. The issue of the wrongly installed washer is quite similar to the wrong washer in that the bolts are allowed to fret.

The key steel in the machined slots around the centre of the bolted joint on the strut is a useful indicator of joint problems when doing visual inspections. Some people advocate using differing colour paint to mark the position of the key, others tack weld the end, especially some of the old timers. If the key shows any sign of movement the joint is moving and the bolts are 1. not doing their job and 2. going to fail.

The bolt is not meant to carry the load via shear stress, that is the job of the machined surfaces.

A related bolting method in heavy earthmoving is to use much longer bolts than seem necessary and fit a machined collar under the head. This increases the spring length and spreads the cyclic load over the full length of the bolt shank. This method is especially popular for anchoring the slew circle on large hydraulic excavators (which are subject to heavy cyclic loads).

Something that needs to be mentioned, is, that bolts fail in service mostly due to fatigue cracking, examination of the broken surfaces will show the "beach marks".

Hardened washers between bolt heads and spot machined surfaces keep the load axial to the bolt shank, reduce torsion locked up in the bolt and prevent creep especially uneven creep. As Randouli states the load needs to be fully axial and at the correct preload.

We changed to washers with bevels to clear the bolt head radius. On the lower bolts, the length was increased with a machined spacer. Every batch of bolts went to the met lab and were tested for hardness and magnafluxed. Parallelism of the mounting surfaces became critical and closely controlled.

You must have been closely involved in the problem and solution.

We are certainly in a similar field. A similar topic is wheel studs. We found that we needed to fit bevelled washers under the original nuts, in a factory approved mod, to be able to get them undone (original nuts have integral washers). Obviously I'm talking about flanged type rims rather than cleated.

Grade 12 bolts are a necessity in this class of equipment, but instalation techniques and quality control is critical. There was a limit of 3 useages on the strut bolts, now it is basically once only. Some trucks that came via Chile I saw recently had the struts removed with the oxy, those bolts won't get reused. Trying to enforce standards when mines use labour hire fitters doesn't always work out.

I'll have to look up where you are in Illinois, I'm off to NZ today so I'll look when I get back.

Welcome to the insanity.

I have not come across this type of failure. However I assume that the bolt goes into a housing (i.e. not a nut) and when the torque is applied, part of the torque produces tension and part goes to damage the washer. There may be sufficient friction in the thread engagement to accumulate torsional strain in the bolt or if the threads allow normal screw motion then the tension will add up during the 'washer damage' process. Washer damage may be acting as a lubricant thus reducing friction under the head and applying all the torque to produce tension.

If the breakage is normal to the axis it would suggest tensile break as in second case or if the breakage is at an angle then it signifies a torsional failure.

When a hardened washer is used all the torque is used to produce tension in the bolt and some lost in friction under the head limiting the total tension and the process of torque application is stopped when the correct torque is registered.

I'm glad I am not alone in trying to understand this phenomenon. I wish the occurance were more recent and that the failed fasteners were at hand. One thing I do know for certain: the bolts and nuts were the same. The only change was the washer.

I appreciate the efforts gentelmen. Thanks.

L.j.

As a machine designer for an automation company we encountered this problem years ago with socket head cap screws clamping soft A2 steel. Someone finally read the fastener manufacturer's catalog where it stated that the grade 8 fasteners must clamp onto hardened steel in order to achieve their maximum clamping force. As others have mentioned the small bearing area under the head of the SHCS will gall and displace metal squandering your torque and dropping your clamping force.

We either had the screw head clamp a hardened plate or made room for HDFW (hardened flat washers).

Try here:

http://www.boltscience.com/pages/strength.htm