# Why didn't the space shuttle use non-foam-shedding external tanks?

Throughout its lifetime, the Space Shuttle's external tank had severe problems with portions of its foam insulation coming apart and falling off, which NASA was never able to fix. The flying fragments of foam frequently caused damage to the orbiter's (rather fragile) heatshield, which was occasionally severe.

Notable examples:

Given the potentially fatal results of ET foam-shedding, why didn't NASA redesign the ET to make it shedproof?

There are several ways they could have done so:

• Using a form of insulation other than spray-on foam (said foam apparently had problems with gasses trapped within it expanding and breaking it apart as the ambient atmospheric pressure dropped during ascent).
• Using a double-walled tank, with an outer structural wall, an inner set of cryogenic LH2/LOX tanks (most likely balloon tanks, to reduce weight), and the space in between filled with insulation (maybe foam, maybe not).
• Another double-walled tank, but this time with a vacuum in between the outer and inner tanks (basically a giant dewar flask), which would not only insulate the contents from the exterior but also save on weight (there would have to be some struts between the two, to support the inner tank; these would presumably have been made of a high-strength, low-thermal-conductivity material, probably some sort of strong ceramic).
• To address your question, and not your proposed solutions, you should read Chapter 6 of the CAIB report "Decision Making at NASA". The topic is too important (and painful) for me to summarize, and there is no better writeup on the failure of shuttle program management. You can find the document here: s3.amazonaws.com/akamai.netstorage/anon.nasa-global/CAIB/… – Organic Marble Mar 9 '18 at 4:59
• A double walled tank with vacuum between the walls would be much to heavy. Strong ceramic would be too brittle. – Uwe Mar 9 '18 at 8:56
• Other insulations would not have prevented the formation of ice on the tank, which would be even worse than the foam for shedding. Literally no rockets that I am aware of are built using the other two techniques you are describing. – Tristan Mar 9 '18 at 16:22
• asbestos tiles could withstand the heat and may be light enough – john Dec 26 '20 at 10:59

Setting aside the political/managerial issues this is about mass. The final version of the tank had a "dry" (empty tanks) mass of about 26.5 tons and a fuelled mass of about 760 tons. The surface area is a roughly 2600 $m^2$ and the total mass of the thermal protection system is just over 2 tons (all masses and dimensions from wikipedia). So any replacement for the foam needs to fit within a mass budget of roughly 1 $kg/m^2$. This rules out most alternatives.