5.5" ATACMS Flight 4

In this graph the acceleration and blue altitude lines are unfiltered while the velocity and green altitude are Kalman filtered.

Once again I am plagued by early deployment. It seems to be a theme with this rocket. Deployment occurred about 9 seconds after motor burnout. Very early for a 14 second (nominal) delay.

I watched the rocket descend under drogue waiting for the Tether to deploy the main. Once again it did not. But this time the fault was not with the Tether or my rigging. If it had been activated in the air I am convinced it would have worked. The problem is that the AltAcc fired its main charge 207 seconds into the flight. How can that be? I hear you ask.

The AltAcc integrates the acceleration and when this integrated value changes from positive to negative this indicates a change from upward velocity to downward. This triggers the apogee charge. The problem on this flight results from the early deployment. This happened well before the AltAcc was ready. Because of the configuration of this rocket, when descending under parachute the altimeter is mostly sitting in the same orientation as it is before launch. As a result it tends to integrate a value of zero. So the integrated value never has a chance to decrease. Once the rocket lands the altimeter should be sideways and then -1G is integrated resulting in deployment a short time later.

Here is a plot of the velocity determined by integrating the acceleration:

While I consider early motor ejection to be the most likely cause of the early deployment, I had to consider the possibility that the shear pins were not sufficient.

The strength of the shear pins I use is about 24 pounds each according to testing at www.rocketmaterials.com. So the required force to shear both is 50 pounds. With a nominal inside diameter of 5.5", this translates to about 2.1 psi. The peak pressure change from ground level on this flight was about 3 psi. So if the vent hole didn't do its job there would be more than enough force to break the pins.

I used a formula intended for computing the time constant of altimeter bays to determine that the time constant of the single 1/8" diameter vent hole was on the order of 6 seconds. This might be too slow so I added a second hole to cut this in half. This also provides some redundancy in case something blocks the hole.

Here is the data in a somewhat more raw form:

AltAcc data file. (binary, 8K)

Processed data (compressed text, 55K)

Kalman filtered data (compressed text, 45K)