Factor of Safety design in space craft structures

[Platypusimon]

Hello everyone,

I would like some advice, clarity and guidance on something that I have been debating with my self for a while in regards to Factor of Safety/Safety Factor and Margins of Safety in aerospace structures design.

It's generally recommended to design parts in aerospace with a FoS of 1.25. However, what is bugging me is the way in which one would like to apply this FoS to their design. I feel like there is two methods, but not standardised choice in which method to use.

Method 1:
Let's say for a metal, you would have Yield Load/Applied Load,
Yield = 500MPa So : Yield/Applied = 2.5
Applied = 200 MPa Given we have a ratio of 2.5, we could say the FoS is 2.5, thus a Safety Factor of 1.25 is well in place in our design.

Method 2 :
Again, let's say a metal and we want a FoS of 1.25, there is the option of doing:
Yield/(Applied*FoS) ---> 500/(200*1.25)= 2
MoS = 2 - 1 = 1 --> MoS = 100%

Which method would you say is the most appropriate? I feel like the 2nd method is safer, however you would lead to design a structure with a higher weight, which is not preferable at all in aerospace applications.

I would strongly like more expert and experiences engineers input on this.

Also, would you say an eigenvalue factor could be considered as an FoS in the case of buckling?

Many thanks in advance.

 

[SWComposites]

a) where do you get the "recommendation" for FoS of 1.25?? FAA requires FoS = 1.5; NASA requires FoS = 1.4;
b) the "FoS" is the value applied to Limit loads to obtain Ultimate loads for analysis or test. So your method 2 appears correct.
c) an eigenvalue "factor" is NOT a FoS; it is typically to account for the unconservative inaccuracy of a eigenvalue buckling analysis.

 

[rb1957]

yeah ... FoS of 1.25 ??

method 2 incorporates the FoS explicitly which is better than method 1 (take shows there is plenty of margin).

I prefer method 2, but would have a hard time saying method 1 is unacceptable.

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[Platypusimon]

I wasn't using FAA or NASA, but standards I had found in Europe. Regardless if 1.25 or 1.4, it's more about the question of the application.

The gain from using method 1, is that if it is under the FoS we are aiming for, then we can have a structure that weighs less, which is primordial in aerospace. Nevertheless, method 2 pleases my "safety side" of things.

I am in awe that there isn't a clear stated international standard method on what is the preferred way to apply FoS. From all the research I have down, everyone seems to do it the way they prefer...

This is why I am asking the question, as a young engineer I am hoping that more experienced and knowledgeable engineers can give me guidance on this question.

 

[rb1957]

nah, sorry ... numbers are important to us. I would be interested in your reference for a 1.25 FoS. It is possibly for a special application (maybe spacecraft) where weight and testing are factors that permit a lower than normal FoS. Normal FoS in aircraft is 1.5.

As I posted, it comes down to what you need to show.

Method 2 incorporates the FoS explicitly in the calculation, and I'd say is preferred.

Method 1 accounts for the FoS narratively which whilst not unacceptable (to me), could be considered insufficient by some reviewers. It depends on the reviewer ! (and their management opinion on the day !!) The challenge with this approach is what MS do you ascribe to the installation ? (> 0.00 ?) The other challenger is clearly understanding the loads used in the analysis. Method 1 uses limit loads, as does Method 2.

So Method 3 incorporates the FoS into the loads, so your loads are ultimate (which is much more common, nearly universal).

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[SWComposites]

OP -
- what specific standard are you refering to?
- what are you specifically designing?
- what is the regulatory agency for this structure?

The standard should be specifying how the FoS is applied.

Your method 2 is by far the most common and accepted approach. But both methods, if designed to 0.0 margin of safety (relative to Limit loads * FoS) will result in the same weight structure.

 

[rb1957]

SWC, I was thinking about that. Is it common to use limit load in your calcs ? I always use ultimate load (PITA to work with aero loads, being limit).

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[SWComposites]

Well, its been done both ways. Often the FEMs are run with Limit loads, then the correct FoS for the load case is applied when writing margins, either in hand calcs or post-processing codes. This helps when one FoS is used for mechanical loads and another FoS is used for thermal loads and the two are combined. Its also easier to check the FEM loads against the documented loads from aero, etc.

 

[rb1957]

sure, never had different FoS to deal with. in that case working with limit is the way to go .

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[rb1957]

as I recall my 1st job (too many years ago) with BAe we didn't separate inertial load and thermal loads ... both had FoS of 1.5 (as I remember).

"Hoffen wir mal, dass alles gut geht !"
General Paulus, Nov 1942, outside Stalingrad after the launch of Operation Uranus.

 

[FEAster_man]

Platypusimon,

Your method 2 is more common notation, but essentially it gives the same answer as method 1.
The thing is that in the method 1 your 2.5 factor is not a FOS, but another value, let's call it factor of safety margin. If you divide it by FOS 1.25 (instead of simply comparing the two numbers) you'll obtain the same margin as in method 2.

Since you mentioned that you are using European standards my guess goes towards ECSS documents. If that's the case, make sure that your "applied load" is correct as there are also other factors to be taken into account.
Also if ECSS are the standards that you are working on, the logic behind MOS calculations is described in ECSS-E-ST-32 for example.