Ot- what is best practice when allowing load safety margins in design?

What is the practice used in allowing some extra margin in fittings when designing for safety?

I was specifying a part strength and a coworker commented that I was failing in the exercise as the goal of 'good engineering' was to "make it strong enough but no stronger".
I am not a engineer but occasionally need to specify a part strength.

So... say a bolt.

I would proceed by citing known loads, adding margin of unknowns to order of maybe 50% additional and call that number 'working load' for component.

The comment was that known load is working load, any additional loads the part sees are taken up by safety margin build into the bolt by nature of the breaking strength being far higher than stated working load.

I am right eh?- "working load" is a approximation of all known and estimated loads and one specifies components to that number?
I guess I would call that the "design load" for the widget?

It seems to me that using up the margin between a the cited working load on a component and breaking strength by designing the part to be used in that margin is not good practice?

Formal practice?- trade practice?
"She looks like it will hold"?

Thanks all

Don't confuse "working load" with yield strength, which is usually the lower number cited when there are two strength/stress levels given for a bolt. In most applications, if a bolt yields, it has failed.

Margins of safety really depend on the application and how well the loads and fatigue cycle are known. Adding 50% to a known load is a margin of safety of 1.5, which is really pretty slim, especially when you start taking fatigue into account. The allowable stresses decrease significantly when the applied load cycles, how much depends on the nature of how that load changes. Common safety factors are in the range of 3 to 5 for a single event, then more analysis needs to happen to understand how fatigue enters the picture.

When in doubt, build it stout, out of material you know a lot about.

Ok- this coworker stated that the "margin of safety" was the difference between "working load" cited on fastener and its breaking strength.

So a fitting provided which has a breaking strength of 5x "working load" is already designing in safety margin for me and that I am over doing it by keeping my estimate of known and unknown loads below the working load provided with the fastener.

So one should proceed by stating known load, multiplying by safety factor, select component with "working load" above that number?

So you will end up with a assembly which is miles away from failure and that is the goal.

Is that clear?- I need more coffee....

If you are designing something against a catastrophic failure, then I would use a fastener breaking strength. I would be hesitant to use a breaking strength, but it is dependent upon your application. Yield would typically be associated with failure of the bolt as Graham stated above. You should design to the proof strength of the fastener.

I don't think that you are being too conservative by keeping your known and unknown loads below the working load of the fastener. Depending upon the use of the bolt, you may have to account for misuse or overloading of the bolt by whomever may end up with the assembly. Though the margin between failure and the working strength is substantial, unforeseen and skewed forces can really play havoc on your design.

I agree with the engineer in that you don't want something that is completely overkill either.

I would continue specifying your bolt size based on your known and assumed unknown loads (design load) being equivalent to the working load of the fastener.

Go get yourself a steel code manual. Different loads require different safety factors along with which theory you use to design your elements. Also keep in mind the yeild strength is the minimum yield. I've seen steel that theoretically fails at 60 ksi fail at above 80 ksi (quite common) So you have "hidden" safety factors in your design also. There is a whole section in the manual that gives you allowable shear and tensile loads. Simple.

Ok thanks guys.

TimRD said:

I agree with the engineer in that you don't want something that is completely overkill either.

Click to expand...

This is not a engineer I am getting this from but he is a guy who has to specify loads in his trade.
His take is- "look at that breaking strength, it's plenty strong enough", where I take the WLL literally and design below the 'limit' and never even bother to look at breaking strength.

Honestly I think he is just picking the sexier number: "look at how 'strong' this is".

The only factor in this instance is cost BTW- lower rated parts are cheaper.

Trboatworks said:

Ok- this coworker stated that the "margin of safety" was the difference between "working load" cited on fastener and its breaking strength.

So a fitting provided which has a breaking strength of 5x "working load" is already designing in safety margin for me and that I am over doing it by keeping my estimate of known and unknown loads below the working load provided with the fastener.

So one should proceed by stating known load, multiplying by safety factor, select component with "working load" above that number?

So you will end up with a assembly which is miles away from failure and that is the goal.

Is that clear?- I need more coffee....

Click to expand...

Yes, what you are doing sounds reasonable. It sounds like your working loads quoted have a safety factor of 5 built in, so in a normal situation, if your known worst case load is below the working load you should be well under a failure scenario. How you figure that "worst case" is important. If this is a critical assembly where its failure could cause injury or loss of life, I would do some more involved analysis beyond tables and rules of thumb. My Machine Design class was taught by an old timer who had designed jet engines and their mountings so I have had a healthy respect for fatigue, safety factors, and worst cases beaten into me.

as others have mentioned safety factor of 4 to 10 is common.
.
4 is minimum and 10 is used if it is critical part where failure would immediately cause a economic loss or bodily injury. basically 10x safety factor is anybody would care the part failed.
.
like SpaceX rocket that blew up because a strut that should have been strong enough at 4x safety failure rating failed and rocket blew up. i believe they are going to pretest every strut now and at least 400% of anticipated load to see if struts are defect free. often metal parts have internal hard to detect defects in the metal. 4x safety factor minimum is from experience over centuries of minimum that works. and it is the minimum and most use a higher safety rating

There is no blanket requirement for safety factors. Safety factors are specified differently for different applications. Usually they are covered by codes or ANSI standards. Some will specify allowable stress. Others will specify a factor of safety with respect to yield strength or with respect to ultimate strength. All will specify the loading conditions to be used in calculation.
Get a copy of the appropriate code or standard and follow it.

In some cases deflection will be the limiting factor. In others it may relate to fatigue life.

It is VERY application specific. A co-driver who does aerospace design once commented that that the FOS in racecars is actually much lower than even military aircraft. And in fact finding and replacing stressed fasteners was and is a routine part of racecar maintence. As in check all the bolts every few hours of running an replace every bolt on the car every season - at a minimum. (Yes, every bolt and fitting on the car....)

Virtually no marine, general transport, or even static machinery application would tolerate that narrow an FOS. Aircraft certainly wouldn't.

And having the strength and ductility and redundency to fail gracefully (e.g. it's noticed in a routine inspection) rather than with great sudden drama ("rudder stuck" while 3 days from any harbor) is key too.