Catastrophic Failure

On June 29, 2015 a Space Ex Falcon 9 lifted off from Cape Canaveral on its way for a resupply of Space Station. 2 minutes and 19 seconds into the flight it exploded spectacularly. Space Ex immediately launched an internal investigation as to the causes of the explosion. Shortly afterwards, Space Ex officials began claiming that had astronauts been on board they would have been safely jettisoned in an escape vehicle. At this same time Space Ex began showing videos of safely executed tests conducted on this jettison system at their facilities. While escape vehicles have been touted by NASA and numerous other space agencies as necessary components of a launch vehicle since the original Challenger explosion many years ago, it is by no means appropriate to think that any such vehicle can protect astronauts during an event such as happened to this Space Ex vehicle.

Escape launch vehicles require advance warning of system problems of a long enough time period so that the escape vehicle itself can be outside the deadly shock wave produced by a large explosion. In the case of this Falcon 9 explosion they had less than a second of data to warn them that there was a problem. This is not a long enough time period to activate the jettison of an escape pod and escape the shock wave. It also assumes that an anomaly such as was seen on Falcon 9, the temporary loss of Helium pressure would have initiated such a sequence. I can assure you that it would not initiate such a sequence. Less than one second of data suggesting that there is an unknown and unknowable blip in pressure would NOT be a cause to abort a mission.

Getting back to the explosion itself, Space Ex eventually concluded that the explosion was caused by a rapid overpressurization of the second stage LOX tank. The first stage of Falcon rockets use an array of Merlin engines. The second stage uses one vacuum tuned engine of the same type with a different injector and nozzle type. Once a vehicle escapes Earth's atmosphere and gravity the power requirements for propulsion drop dramatically. One engine is sufficient to propel the vehicle at this stage but it has to be an engine that is more efficient operating in the perfect vacuum of space. NASA utilizes Hydrogen/LOX engines in this atmosphere because of the increased specific impulse energy over a RP1/LOX engine. For simplicity and the elimination of different fuel storage and handling techniques Space Ex early on opted for RP1/LOX engines for both stages of their vehicles. It is a trade off between efficiency and weight, which is at the root of all such decisions concerning space vehicles.

In order to support combustion of a powerful rocket engine you must have a very fast and efficient pumping system to provide it with the massive amounts of fuel required to produce the high thrust requirements of such a vehicle. Turbopumps have long been a source of problems in this chain. Turbopumps on LOX systems are specifically very problematic because of a couple of inherent features. A turbopump runs at extremely high rates of speed with very close tolerances. As I mentioned before, LOX systems preclude the use of lubricants because of the incompatability of LOX to hydrocarbons. Therefore, you wind up with tight tolerances at extremely cold temperatures that have to self lubricate with the LOX itself. Every rocket manufacturer has struggled with this architecture. As I will explain later, it was this same type of system that was at the heart of the Antarres explosion earlier in this same year.

One of the ways of alleviating this problem is to run the initial tank pressures at high pressures to begin with. This becomes somewhat problematic in smaller tanks as it requires a finely tuned pressurization system to balance the ullage (gas pocket) that sits on top of the liquid with the rapidly depleting liquid in the tank when the engine is firing. Falcon 9 uses gaseous helium to provide the pressurization in their LOX tank on the second stage. Compressed helium is stored in several small pressure bottles inside the LOX tank itself at 5500 PSIG.

In order to give a relative idea of what this type of pressure is capable of think of the movie Jaws where in the last scene the shark is blown up by shooting a rifle into a scuba tank inside his mouth. Scuba tanks are pressurized with breathing air to a little less than 3000 PSIG. As you might imagine such high pressures require a very thick and heavy tank wall so that the vessel itself won't rupture inadvertently from this internal force. As I have mentioned already, the crucial tradeoff in space flight is always weight vs. safety. Instead of 4-1 safety factors such as a scuba tank on earth requires, space vehicles use 1.4 - 1.5 to one as a basic safety requirement. Instead of a tank that operates at 5500 PSIG being required to be designed to withstand 22000 PSIG it is designed to withstand 7700 PSIG.

Space Ex is also doing several other things with this system to increase efficiency. The first thing they are doing is super cooling the helium tank and fill gas by storing it inside the LOX tank itself. This is not a new discovery as NASA has been doing this for quite some time on many different vehicles. By putting the pressurization tank inside the LOX itself it is possible to have a smaller pressurization tank. A tank at 5500 PSIG Helium in ambient temperature will contain much less actual usable gas than a tank containing 5500 PSIG at -279 Degrees F. The Helium molecules become denser and can be more efficiently compressed so that more gas is available to pressurize the tank.

Falcon 9's system utilizes the engine itself to heat the helium and rapidly expand it before returning it to the LOX tank ullage to pressurize the LOX system. The helium is released through chambers in the engine as it burns which rapidly expands the helium before it is returned to the LOX tank to be used as a pressurization source to force the LOX out of the tank and into the engine. As you can imagine, the control system needed to tightly control the ullage at a constant pressure utilizing this chain of events is quite complicated.

Naturally, Falcon 9's LOX system is as full as possible prior to flight. A large initial ullage would require a large tank and more weight so at launch the ullage in the LOX tank is very small. This effectively means that relatively small loss of control of this pressurization system will result in large pressure fluctuations in the LOX tank during the initial use of this system. Rapid pressure rise from uncontrolled release of the Helium in these tanks has resulted in the loss of both Falcon rockets that have exploded so far. We know it did with the first Falcon explosion and all evidence points in the direction that it did on this last one as well.

Before we go any further on this concept let me get back to the other thing that Space Ex is doing with their helium pressurization system. Remember the analogy of the scuba tank for a moment. Not only is Space Ex using a 1.4-1 safety factor on the design of the tanks to be used in but they are also using some relatively new technology to build the tanks themselves. Composite Overwrap Pressure Vessels (COPV's) have been under development for some time now. We have built and tested to failure many such vessels where I work so I am not exactly unfamiliar with the concept.

A COPV can be manufactured in several different ways. Most of them utilize a lightweight inner aluminum tank that is rated at a very low pressure. It is thin walled aluminum and extremely lightweight in comparison with an equal stainless steel tank of the same size. This thin walled aluminum tank is then wrapped with carbon fiber, usually by an automated computer controlled lathe in numerous layers of thin carbon wires that are glued in place as the wrap process takes place. The resultant tank is then tested for pressure rating. Design processes and computer programs to produce these tanks are complicated to say the least.

The end result is a pressure vessel that is vastly lighter, which makes it ideal for usage on space flight vehicles. We have tested design and process on these tanks for many years in our shop. Properly done and controlled such processes produce reliable and predictable results. We have tested series of tanks to failure at Liquid Nitrogen Temperatures (-320 degrees F) and even up to Liquid Helium Temperatures (-452 Degrees F). NASA has entertained the idea of using these tanks just as Space Ex currently does in LOX tanks. They have even considered doing the same thing in Liquid Hydrogen tanks which is why we tested at even colder temperatures.

The key to the successful production of such tanks is process control. In other words, careful control of materials and manufacturing process produces consistent results in the actual pressure these tanks will withstand. Any loss of control of either of these variables produces disastrous results which is why NASA is still reluctant to commit to using COPV's inside LOX and LH2 tanks.

There are at least two other dangers associated with COPV's and their usage. Each pressure cycle effectively stretches the composite overwrap strands to a certain extent. Therefore, COPV's are listed for usage by the number of pressure cycles they see as a matter of course. Each pressure cycle is noted and logged and after a sufficient number of cycles the tank is no longer rated for use as a pressure vessel. By itself, this is not a problem for the usage of COPV's on space vehicles as they will see a relatively small number of usages anyway.

The last part of this issue is that COPV's are extremely fragile as far as handling and usage in comparison with metal tanks. Since the thin composite strands are effectively a very large chain reaction system, any external damage to a small strand effectively makes it the weakest link in a chain and we all know the adage about the weakest link in any chain being the source of its strength. Inadvertent bumping into sharp objects or dropping of objects on COPV's can cause catastrophic failures accordingly.