QUESTION: Static Load Test of a Cut-Out of Large Cylindrical Structure

[JulesVerne77]

can anyone help me..
we have to arrange a static load test for a large cylindrical structure( R=2000, L=1000,t=30mm), but we are only able to test a cut-out of the cylinder.. in a fe-model we determined the internal force at this location (axial compressive forces, radial an circumferential forces, tangentail bending moment), we try to apply these forces on the surfaces of the shell (cut-out).. Unfortunately we are not able to apply the circumferential forces (hoop).. only the axial compression, radial forces and the bending moments..

now, my question.. what effects/differences can we expect due to this assumption.. is it possible to estimate these effects.. how will the axial, radial stresses be changed..

we would be glad, if anyone could help us.. thanks..

 

[prex]

If you don't test the real structure with real loads, so what's the meaning of the test?
Anyway to give a sensible answer we need to know everything about your structure in both the real condition and in the test condition: type and distribution of load, geometry, supports, materials, the goal of the test (just a normal integrity test or...). prex

http://www.xcalcs.com

Online tools for structural design

 

[JulesVerne77]

@prex,
the background is a global static load test to qualify a cylindrical cylinder (material= AL-Honeycomb core (xxx) and CFK facesheets).. the cylinder is fixed to the adjacent upper and lower structure by rivets in a al-bracket (aaa)...

¦xxx¦ Sandwich
¦xxx¦
¦¦xxx¦¦
¦¦xxx¦¦
<¦¦---¦¦> Rivet
¦¦xxx¦¦
¦ ¦
¦aaaaa¦
¦aaaaa¦ AL-Bracket
¦aaaaa¦

the problem that we can^t use the lower structure.. so the stresses at the lower interface will not be representative..
To qualify the lower interface we have to qualify the lower interface in a local static load test.. as described above, the sample is only a cut-out of the cylinder... the main loads are compressive loads in axial direction, tangential bending moment and the hoop stress(compressive)..

we are not able to apply the hoop stress.. but we assume that the compressive hoop stress lead to higher margins of the structure, so we have some safety..

but it is difficult to estimate the effects, which exclusive are based on the absent hoop stress.. we`d like to estimate the difference of the stress contribution with and without the hoop stress, to qualify the structure in this local load test..

jules

 

[prex]

Again I would ask for more precisions to give an answer:
1)The test is intended at qualifying exactly what: the connection to the lower structure? to the upper structure? both? the stability of the sandwich to axial compressive stress? the same to circumferential compressive stress? both?
2)How are loads applied in the real structure? Weight resting on top of upper structure, lower structure supported onto ground? Are there any loads (local or distributed over surface) directly applied to the honeycomb?
3)Is the stiffness of the lower or upper structure (or both) sufficient to assume that the honeycomb is kept cylindrical under load?
4)Are there longitudinal joints in the sandwich and how much are they rigid with respect to the rest of the sandwich?
As a first comment I would say that if your concerns are about the riveting system, well of course a local small test would suffice, but if what you want to insure is the stability against compressive stresses of the sandwich, then your partial test wouldn't be much representative. It could be used to assess the resistance to buckling of the sandwich (that I assume you already assessed by analysis), but you would be required to extend the results to the real thing by analysis, so that, if that does matter, you couldn't define your test as a 'full size load test'.

prex

http://www.xcalcs.com

Online tools for structural design