Вы находитесь на странице: 1из 1

the "extreme" response calculated is typi- problems with an approach that is nearly as

cally statistically appropriate only for simple: First, we identify the potential failure
assessing the acoustic environment sepa- modes of concern in the test article. Second, we
rately (e.g., for an acoustic test). There is directly calculate the parameters that relate
little chance such a high load would occur to those failure modes. (Calculate stresses and
precisely when the transient load peaks.1 bolt loads in the loads analysis in addition to
Also, the acoustic environment for most accelerations.) Finally, we define static load
launch vehicles does not peak when tran- cases that we predict will cause the critical
sients are highest. stresses and bolt loads to be at or slightly
higher than those for design, subject to the
Downloaded by KUNGLIGA TEKNISKA HOGSKOLEN KTH on December 1, 2015 | http://arc.aiaa.org | DOI: 10.2514/6.1997-882

appropriate test factor.


Load factors: nx = 23 In the above example, the failure modes of
ny = 15 concern might be bolt tension/shear interaction,
Box weight = 35 Ib nz = 20 insert pull-out, and panel bending. In this case,
we would calculate design limit bolt loads and
Px = 23(35) = 81 Olb panel bending stresses, multiply them by the
test factor (say, 1.25), and then devise one or
more load cases that are predicted to cause
these target parameters. We do this by gener-
ating a load-transformation matrix (LTM) that
relates the selected parameters to the applied
Pv = 15(35) = 530 Ib loads. Figure 3 shows how to generate an LTM
for computing bolt loads in terms of three forces
applied to the box loading head, assuming the
mounting interface is rigid. (In an actual prob-
lem such as this, to account for indeterminate
Fig. 2. Example Structure that will be load paths, we would use finite-element analy-
Statically Tested. The objective is to sis to generate an LTM. For the purpose of this
verify the panel's strength for the design example, we will assume the LTM equations in
limit loads shown. Fig. 3 are accurate.)
To complete this example, let's assume our
Much of the peak box acceleration calcu- goal is to devise a static load case that will
lated for acoustics, which typically in- qualify the interface for peak bolt tension. By
cludes vibration up to 2000 Hz, may not planning ahead, in the loads analysis used to
load the mounting panel nearly as establish design loads, we calculate tensile
severely as uniform acceleration. Higher forces in the four interface bolts. We do this
order box modes of vibration, in which either by using another LTM or by constructing
parts of the box are accelerating in differ- the model to enable direct force recovery from
ent directions at the same time, may be the finite-element analysis, such as with
contributing to the total response. In this spring elements that represent the bolts.
regard, static testing is overly conserva- Assume we also calculate response accelerations
tive. at the four outboard corners of the box. The
following limit loads result:
During launch, the panel may be stressed
by other loads than those introduced by Peak accelerations at any corner:
the box. Distortions of the boundary struc- *max =23 g
ture may cause significant stresses. Or the >/max=15 g
panel itself may respond significantly to
Zmax=20g
acoustic pressure. Thus, again, the
planned static test may not be adequate. Peak tension in any bolt:
Clearly, we cannot design an effective p, max =190 Ib
static test based solely on box accelerations. From the LTM in Fig. 3, we can generate a
Instead, we can address each of the above predicted 190 Ib tension in any one fastener by

American Institute of Aeronautics and Astronautics