Analysis and Design of a "Catching" Device
Background
The Shock Test has to be performed in the frame of a PayLoad Fairing (PLF) Validation Campaign. The Shock occurs when the PLF is ejected from the rocket (see Figure 1).
The test is done under gravity acceleration and the ejection is executed by a Spring Pack assembly able to accelerate the PLF at 2g.
The target of the test is to measure the shock and the trajectory of the PLF.
The Spring Pack is released by pyro, commanded with an external pulse; in the real operative condition the PLF will be split and drift away, but in the test execution no split occurs and the PLF has to be catch and hold.
The Catching System is responsible to catch the PLF avoiding to fall back on the basis, but it also shall not interfere with the measurement performed (i.e. it has to intervene at the end of the PLF ascending trajectory).
The Challenge
The working principle of the Catching System is a falling mass which acts as counter-mass.
The falling of the counter-mass shall be triggered by the physical movement of the PLF and not by the electrical command sent to the ejection system.
This is done to avoid that the mass will drag the payload in case of failure of the pyro on the ejection mechanism (pulse sent but not executed).
Moreover it shall decelerate until a complete stop of the PLF.
In the sketches in Figures 2 and 3 the PLF is represented by the Mass Mo, while Mass M1 is the counter mass.
Some descritpion of what will happen during the test is also given in Figure 2 and 3.
Figure 4 shows in a graphical way the different phases of the test
- T0 = The ejection is commanded
- T1 = The switch is triggered
- T2 = The counter mass starts to fall
- T3 = The PLF reaches the maximum height before falling down
- T4’ = moment in which the falling counter mass starts to actively pull the PLF
- T4’’= moment in which the counter mass “drags” the PLF
In order to give a better idea on how much challenging the challange was, Table 1 reports some data concerning the system and the masses taking part to the game.
Considering that all the test is lasting 200 ms, it is clear that there was no margin for errors in the simulaltions!
CG CAE was chosen as a partner to support the design with all the simulations required exactly for this reason: our vaste experience in Finite Element Analyses in different high tech environments gave us the confidence to profitably take on board this challenge.
Finally, Figure 5 contains some images of the schematic of the system, together with some of the parts that were designed and calculated.
The FE Models
Because it was clear since the beginning of the activity that a big number of simulations would have been necessary, the decision to build a simplified though realistic model was taken.
Therefore a plane strain model (see Figure 6) was considered in order to be able to run many simulations in a reasonable time.
The phenomenon under investigation is a high speed non linear dynamic one, and therefore an Explicit solution approach could have been considered.
Nonetheless the Implicit approach was used, to get rid of any possible doubt concerning the numerical quality of the results (see Chapter 14 of the book Computational Structural Engineering for more information about Explict and Implicit approaches).
Figures 7 and 8 are giving more details on how some parameters (like the stiffness of the PLF Interface Structure) of the plane strain model were tuned.
Many parameters are playing a big role in the behaviour of the system.
The most important is for sure the stiffness of all the elements involved in the chain connecting the two masses.
Therefore FE models were used to size those elements to maximize the stiffness.
Nevertheless the rope is a commercial item and the only way to get the maximum stiffness out of it has been to make it as short as possible and to pre-stretch it.
Pretension of the rope was also relevant: a too high value would have significantly affected the PLF behaviour, whereas a too low one could have led the rope to “jump” out from the pulleys.
The time delay with which the counter mass is released is another determining factor: too late would have meant having the PLF dropping to close to the spring pack; too early would have affected the PLF trajectory.
Some results
As already mentioned, many anlyses were run and many results were obtained.
The most relevant are those generating graphs, like the few reported in Figure 9, because they give a direct indication on what is happening to the system: we remind that the travel of the Spring Pack is just 8 mm over a height of the PLF equal to 14000 mm!
Figure 9
For sure the most suggestive pictures are those with some animations coming from the simulation, such those contained in Figure 10
Figure 10
Figure 11 is showing some results on some of the components that were sized to withstand the forces acting on them.
Morevoer those Finite Element Models were used to extract, as mentioned, some values to be used in the 2D Plane Strain model.
After all the components were sized and parameter variations were considered, additional analyses were done for other possible deviations:
1. Mass PLF variation (lower and higher)
2. CoG position variation (mainly in the lateral direction)
3. Unbalanced ejecting force (possible failure of one or more springs of the spring pack)
Each of these parameters, plus others, was checked and measures to mitigate the related collateral effects were put in place.
As an example, the lateral deviation of the PLF’s CoG was resulting in a rotation; even if that rotation was small, given the size of the PLF the corners were moving quite a lot and therefore some bump stops have been placed to prevent it to float around once hanged to the rope.
Conclusions
The test was actually performed in the USA, where a big facility to host a so tall test jig exists.
But before performing the actual test (that could not fail!) many pre-tests where done to check all the devices to be used, including the air damper
The videos in Figure 12 show a couple of them.
Figure 12
