Henry Spencer, computer programmer, spacecraft engineer and amateur space historian
NASA's new Ares I launcher has problems: it's too small, it has various technical snags like excessive vibration, and its cost and schedule estimates are worsening steadily. The Augustine Commission is currently investigating whether NASA is really on the right track with it. Now a new problem has surfaced, and it's bad, possibly the death knell for Ares I.
Ares I's first stage is a derivative of the shuttle's Solid Rocket Booster (SRB). This was considered acceptable because of the SRB's supposedly-excellent safety record (it's only killed one crew...), and because the agency's future Orion crew capsule, unlike the shuttle orbiter, would have an escape system to pull it clear of a malfunctioning rocket.
Unfortunately, it turns out that if an abort occurs any time in a sizable part of Ares I's SRB burn, then even if the escape system works perfectly, the capsule will crash and the crew will die.
Past experience, on the shuttle and the Titan rockets, suggests that large multi-segment solid rockets have a probability of failure of 0.5 to 1 per cent. Since a failure would be unsurvivable during about a third of the SRB burn time, that puts the chance of losing a crew on each Ares I launch at 0.2 to 0.3 per cent. This is a far higher risk than NASA's modern rules permit.
more:
http://www.newscientist.com/blogs/shortsharpscience/2009/07/death-knell-for-nasas-ares-roc.html#more