A Bridge Too Far....

A bridge too far is simply an act of overreach; a step too far in advance of capability. In psychological terms it is a symptom of hubris or an over inflated sense of achievement that leads to disaster. At this point at least, it appears that Elon Musk and Space Ex are victims of their own bridge too far. Success in returning rockets and relanding them have been fantastic and extremely impressive but they seem to have led to an attitude the skips over the careful understanding of process in search of further achievement.

I will once again point out that all of the data is not in yet on what happened to cause the Falcon 9 to explode on the pad at Cape Canaveral during loading. While it appears that a COPV in the second stage LOX tank exploded and caused the destruction of the rocket itself; the investigation is ongoing.

In my last post, I went into some detail about the fragile nature of COPV's as far as mechanical external damage is concerned. I can't adequately express how important to the health of a COPV that basic handing integrity happens to be. While COPV's can and have been designed for massive overpressurization limits because of the lack of weight associated with more composite wraps, they are also extremely vulnerable to external damage that might immediately seem inconsequential, but can quickly lead to the failure of the COPV.

I have also pointed out that Space Ex didn't seem adequately concerned with handling of COPV's, both with the mounting methods they use to put them inside the LOX tanks and by the lack of understanding of some of their personnel that they are not meant to be stood on. For whatever reason neither the person standing on the tanks, nor the person distributing pictures documenting this act seemed aware that this might be a problem which of course leads to the question of how many other people in the handling chain for these tanks at Space Ex were similarly unaware.

While this is troubling, it is hardly proof that it caused either or both of the recent Falcon 9 explosions. As I pointed out in my last post, the sequence of events if that is what happened does seem to match the available data. There are a lot of other possibilities however. I will try to briefly explain some other things that could have quite possibly caused an inadvertant rupture of one of these same COPV's.

Let me go over another possible scenario that could have led to the failure of one of these COPV's well below its rated pressure. There are a couple of physics theories that have to be explained in order to understand the following sequence of events. The first one is something called the Joule-Thomsen effect. What Joule-Thomsen says is that expanding a compressed gas across an orifice predictably changes its temperature. This is largely the basic theory that led to the discovery of modern day air conditioning. Most gases at or near ambient temperature will tend to cool as they expand across an orifice. Therefore, taking a compressed gas such as Freon and expanding it across an orifice or valve will chill that gas quite dramatically. This cooling effect is exactly how a modern day air conditioner works. Compressed freon is expanded across an orifice into a cold tubing bundle. If air is forced across this bundle by means of a fan, the air itself is chilled accordingly. Air conditioners are a closed loop system. This means that the same Freon gas is later collected and forced through a compressor which is used to pressurize it again so that it can repeat this basic process again.

There is a rather odd anomaly surrounding three gases when it comes to the Joule-Thomsen effect. Hydrogen, Helium, and Neon have what is called a negative Joule-Thomsen effect under certain conditions. In other words, they can produce heat when expanded across an orifice, dependent on starting temperature of the gas and pressure. In the case of Helium space vehicle pressurization systems tend to cross over and through this barrier as a normal event. In other words, part of the time that you are transferring helium through a valve or orifice it will get colder and then it will go through a stage where it produces heat and then back to colder. It is predictable, if not easily calculated but it is something that should always be taken into effect when pressure rating helium and hydrogen systems. Higher heat inevitably leads to lower pressure ratings for containment vessels, valves, and pipelines in such systems.

The other side of this same process then has an effect. It is called adiabatic heating. Adiabatic heating is simply heat of work that is added to the Freon as it is stepped up to a higher pressure again by the compressor. This heat is usually dissipated in large coils that are exposed to circulating air which tends to cool the gas by the time it gets back to the orifice to be expanded again.

Another example of Adiabatic heating is a diesel engine. Diesel engines don't require spark plugs to light the fuel in each cylinder because they relay on the adiabatic heating process that occurs when the inlet valve is closed and the piston is driven at a high rate of speed inside the cylinder to compress the air/diesel mixture of gas in the cylinder. The adiabatic heating combined with the piston still being hot from the last stroke is high enough to light this mixture causing an explosion which forces the piston back down. This in turn causes the main crank to turn and push another piston up with the same result.

Adiabatic heating can be both large and fast, dependent on the gas being compressed and its volume. It is therefore also an ignition mechanism in fluid systems that has to be carefully considered when transferring gases at high pressures and volumes. Extreme heat weakens pressure ratings because the metals themselves lose strength and become more ductile when heated. This difference in tensile strength can be very large, depending on material, rate of heating, and purity or metallurgical control of manufacturing process.

As I discussed earlier one of the relative oddities of COPV's is that they will actually contain higher pressures at cryogen temperatures. This is because the extreme cold at tends to shrink the composite fibers more than the internal aluminum tank. This same quality reverses as a COPV is heated. The aluminum tank tends to expand to stress the carbon fibers more, putting them in a preload condition that limits the expansion they can take as the inner tank starts to try to expand with increases in internal pressure. If a COPV is slowly and uniformly heated this effect is lessened. If it is heated in a non-uniform matter it increases.

When we have used COPV's in testing Shuttle systems we routinely monitored internal temperatures in them to guard against this effect. Very minor increases in temperature inside the tank while the outside carbon fibers remain at the same temperature were deemed problematic enough that we had cutoffs that stopped the pressurization process automatically if we gained over 120 degrees F on the inside of the COPV. I know from experience that this was easily done with a very small volume pnuematic system so that we would take a very long time to charge the COPV's to operating temperature to do our tests.

These numbers were based upon a basic 50 degree F differential between the inner part of the COPV and the outer carbon fibers since we were operating at a roughly 70 degree ambient environment. It is routine practice to monitor internal temperatures on COPV's as they are pressurized so that you don't effectively make them able to withstand less pressure. From what I have been able to gather, Space Ex also monitors inlet pressure while filling the COPV's on Falcon 9. What I don't know is how they are monitoring this temperature.

Let me explain why this is also important. In the first place, the most common temperature measurements are made by utilizing a thermocouple which is simply a connection of two dissimilar metals. These dissimilar metals produce a voltage at their junction that changes with temperature. By monitoring these changes it is possible to accurately read this voltage and infer a temperature.

There are other instruments that are used to make such measurements but this is the most common and least expensive as well as being the fastest at monitoring temperature changes so I am somewhat giving Space Ex the benefit of the doubt in assuming they have an accurate measure of internal temperature on the COPV's as they fill them. When I say fast, I mean the amount of time it takes to actually reflect a change in temperature. In the data aquisition world temperatures are amongst the slowest measurements for the simple reason that even the fastest instruments for temperature measurement are rather slow. A step response of 200 milliseconds on the fastest thermocouple would be very good. In other words, dependent on instrument and data rate it is very hard to see rapid temperature changes.

Since Space Ex is filling several COPV's at the same time I am sure they are utilizing a temperature sensor on a tube instead of individual temperature sensors inside each COPV. This matters because the relatively large volume of a COPV inside of a tank filled with LOX (-297 Degrees F) will necessarily read a much lower temperature as the small area inside the tube will be more readily chilled by the LOX than the volume inside the COPV. In essence the temperature monitoring instrument they are using on the COPV's as they fill them is relatively slow to react to changes and physically masked by the physics of its mechanical location.

Let's also keep in mind that the physical environment of the COPV's is greatly exacerbating the relative weakening of its ability to hold high pressure gas. The COPV is seeing at least -297 Degrees F on its carbon fiber exterior at the same time that helium being heated by adiabatic heating and negative Joule-Thomsen effect is being introduced into its interior. We are probably talking a 400 degree F temperature differential that it is routinely seeing on Space Ex's rapid fill propellant transfer which will necessarily vastly lower its pressure rating. Depending on how high this differential actually is, it could rather easily exceed the safety margin of the COPV's pressure rating.

This is the part where we begin to get into the "Bridge too Far" analogy. Space Ex and Elon Musk have their eye on much bigger things than the resupply of Space Station that Falcon 9 is specifically tasked with performing. Musk has made no secret that his eventual goal is the colonization of Mars and they are as I write this developing heavier launch bigger vehicles to reach this goal. It is this ultimate goal that is causing them to push the envelope so hard on Falcon 9. There has been a lot of publicity about reusable rockets and a very public "competition" between Space Ex and Blue Horizon to return rockets and land them successfully in order to keep costs low on launching payloads.

This requires extra fuel and oxidizers to be loaded onto Falcon 9 in order to have the requisite power to control its landing process manuevers. To their great credit both Space Ex and Blue Horizon have managed to be very successful in recovering their rockets with such manuevers. Since this is not what Falcon 9 was originally designed or contracted to do it is has been a little problematic for Space Ex to find ways to get the extra fuel and oxidizer on board to carry out these manuevers and still deliver its payload to Space Station.

One of the ways they have managed to do this is to steadily increase the supercooled effect of the LOX they use on Falcon 9. Supercooling of LOX is not a new idea either, as NASA has used this idea to increase turbopump efficiency for quite a few years. What it involves is decreasing the temperature of LOX away from its boiling point and towards it freezing point. In other words, the colder they can get their LOX before launch, the more energy it will contain. This effectively gives them the margin they need to carry out the extra manuevers necessary to safely land the rockets on their return descent to Earth.

One of the main factors for having super cooled LOX is to load it quickly and to make sure it does not sit in the LOX tank for long before launch. The longer it sits in the tank the more it tends to stratify which leads to less efficient burns and more LOX expended on ascent. Space Ex has been constantly speeding up its rapid load process for quite some time now. This means that massive amounts of LOX and RP1 are transferred to the vehicle just 30 minutes before launch. It also means that Space Ex must launch immediately after loading so that all other checkouts and final preps must occur after loading propellants.

The latest load of propellants as a dry run test was the most rapid load they have yet attempted. This effectively means that the LOX in the tank was colder than it has ever been on a propellant load. It also means that the Helium fill which must occur after LOX is loaded to take advantage of the increased density that putting the COPV's in the LOX tank in the first place was necessitated by. In their every increasing haste to cut the time that LOX sits in the vehicle I am sure they are also accelerating the rate at which they fill their COPV's which also increases the heat load by adiabatic heating and negative Joule-Thomsen effect.

In other words, for reasons completely outside of their actual manifest task, Space Ex has neccessarily vastly increased the temperature differential their COPV's see on propellant load. Is this relationship the key to the latest loss of a Falcon 9 on the pad at Cape Canaveral? I guess time will tell if this is actually a bridge too far or just another bump in the road for Musks' frenzied path to Mars.