NHPR got my attention this morning when it ran a piece about the FDA bashing the anti cholesterol drug Simvastatin. I happen to be on Simvastatin, so that story struck right home. The FDA spokeman was strongly against use of Simvastatin and said it should have been taken off the market years ago. There are many other drugs that are better. Well, yes there are, namely Lipitor. Trouble is, Lipitor costs $3 a pill, Simvastatin costs $0.13 a pill at Walmart.
So I googled to find out what is going on. FDA ran a big study, 6000 people using the big 80 mg dose of Simvastatin and 6000 people using the smaller 20 mg dose. Less than 1 percent of the 80 Mg users came down with a rare muscle ailment that I never heard of. Less than 0.1 percent of users of the smaller 20 mg dose suffered from the same ailment. Well, that's statistically significant. The FDA gave no information linking the difference in risk to the drug, as opposed to underlying conditions in the patient. Patients taking the 80mg dose are doing it 'cause their cholesterol counts were higher and needed a stronger dose of Simvastatin to control it. Could be that patients with higher cholesterol counts are more vulnerable to the rare muscle ailment, but we will ignore that.
Checking my medicine cabinet I find I'm taking the 40 mg dose, not the 80 mg dose, so I'm OK there. Plus, rare muscle ailment usually strikes within a year of starting Simvastatin and I've been on it for longer than that.
Thursday, June 9, 2011
Wednesday, June 8, 2011
CEO of Government Motors calls for gas tax hike
Story is here. Aren't you glad to have your tax dollars bail out this company? Also note the comments on this article, most of them are in favor of the proposed gas tax hike.
Pilot Error
Last year Air France flight 447 disappeared over the south Atlantic. No distress calls, just gone. After a year long search of deep sea bottom, the wreck was located and the flight recorders were recovered from 10,000 feet. They must build 'em right cause after a year on the bottom of the ocean, they still work.
The flight recorders reveal that the aircraft, cruising at 38 thousand feet, had the autopilot suddenly trip off line, leaving the pilot[s] to hand fly the aircraft. This they failed to do. The aircraft went nose up and stalled. And stayed stalled, the pilots never pushed the stick forward to put the nose down, gain airspeed and fly out of the stall. The stall warning horn went off, but the aircrew failed to put the stick forward. The aircraft, stalled, fell from 38 thousand feet (seven miles high) and hit the sea.
This is really hard to believe, 'cause everyone knows about stalls, and everyone knows you push the stick forward to recover from a stall. There were three pilots on the flight deck, and not one of them came up with the right answer.
Contributing factor to the accident, the pitot tube iced up, which caused the indicated airspeed to drop to 60 knots. This is why the autopilot dropped off line, it is smart enough to know that the plane won't fly at 60 knots, something must be wrong, and it wanted the pilot to take over. Unfortunately the pilots could not fly the aircraft either.
The flight recorders reveal that the aircraft, cruising at 38 thousand feet, had the autopilot suddenly trip off line, leaving the pilot[s] to hand fly the aircraft. This they failed to do. The aircraft went nose up and stalled. And stayed stalled, the pilots never pushed the stick forward to put the nose down, gain airspeed and fly out of the stall. The stall warning horn went off, but the aircrew failed to put the stick forward. The aircraft, stalled, fell from 38 thousand feet (seven miles high) and hit the sea.
This is really hard to believe, 'cause everyone knows about stalls, and everyone knows you push the stick forward to recover from a stall. There were three pilots on the flight deck, and not one of them came up with the right answer.
Contributing factor to the accident, the pitot tube iced up, which caused the indicated airspeed to drop to 60 knots. This is why the autopilot dropped off line, it is smart enough to know that the plane won't fly at 60 knots, something must be wrong, and it wanted the pilot to take over. Unfortunately the pilots could not fly the aircraft either.
Tuesday, June 7, 2011
Cybertage, large scale
Aviation Week has a couple of dramatic pictures. The first shows the new power room of the Shushenskaya hydro electric plant. Huge, slightly curved room, big windows, spotless concrete floor with eight turbines neatly set out.
Second photo shows the same room after the accident, the place is wrecked, concrete floor torn apart, turbines and generators ripped apart and hurled into corners.
How did this happen? A plant employee used a network to remotely and accidentally activate an unused turbine with a few errant keystrokes. This created a water hammer that flooded and then destroyed the plant and killed dozens of workers. This happened back in 2009 in Siberia.
The obvious conclusion is that cyber hackers could make the same thing happen deliberately.
Which is why control of electric power generators should NEVER be done over the public internet, and should NEVER rely upon Windows computers.
Second photo shows the same room after the accident, the place is wrecked, concrete floor torn apart, turbines and generators ripped apart and hurled into corners.
How did this happen? A plant employee used a network to remotely and accidentally activate an unused turbine with a few errant keystrokes. This created a water hammer that flooded and then destroyed the plant and killed dozens of workers. This happened back in 2009 in Siberia.
The obvious conclusion is that cyber hackers could make the same thing happen deliberately.
Which is why control of electric power generators should NEVER be done over the public internet, and should NEVER rely upon Windows computers.
Ads slowing the net
I click on a web site, and wait, and wait. Firefox flashes up a running display of all the websites visited on the way to the blog I clicked on. All ad sites. I gotta wait for all the ads to load before I get to see the content I cared about.
There is an opening here, for a web site to claim faster than the average site. All it has to do is display content first and ads second.
There is an opening here, for a web site to claim faster than the average site. All it has to do is display content first and ads second.
Monday, June 6, 2011
Development cost at SpaceX one tenth of NASA.
Aviation Week says SpaceX developed and launched the Falcon 9 heavy lift rocket booster for $390 million. For comparison NASA used the NASA-Air Force Cost Model computer program to estimate the same job and found the computer estimated cost to be $4000 million, ten times as much. The cost savings are attributed to SpaceX program management technique.
NASA administrator Charles Bolden said "They don't spread things all over the country the way that NASA and defense contractors tend to do. They're very focused in two locations in the country. They bring everything in-house. They have no subcontractors, so everything comes to them."
NASA post Shuttle plans are up in the air. They ought to just purchase the Falcon 9 off-the-shelf to boost crew and cargo to the International Space Station. Congress wants NASA to develop their own heavy lift booster to keep the vast Shuttle workforce employed. The Obama administration wanted to drop the NASA heavy lifter development and concentrate on deep space missions. So far the three sides have not agreed on a policy, so things are just drifting. Meanwhile the US is paying the Russians something like $25 million a ticket to take US astronauts up to the International Space Station.
NASA administrator Charles Bolden said "They don't spread things all over the country the way that NASA and defense contractors tend to do. They're very focused in two locations in the country. They bring everything in-house. They have no subcontractors, so everything comes to them."
NASA post Shuttle plans are up in the air. They ought to just purchase the Falcon 9 off-the-shelf to boost crew and cargo to the International Space Station. Congress wants NASA to develop their own heavy lift booster to keep the vast Shuttle workforce employed. The Obama administration wanted to drop the NASA heavy lifter development and concentrate on deep space missions. So far the three sides have not agreed on a policy, so things are just drifting. Meanwhile the US is paying the Russians something like $25 million a ticket to take US astronauts up to the International Space Station.
Friday, June 3, 2011
Robert Gates gives a farewell address
Gates, the out going defense secretary made a few remarks, reported upon
here
He bemoaned the Pentagon procurement system but without offering either an explanation of what is wrong with it, let alone how to fix it. This is unfortunate, so I will attempt to lay out the real procurement problem, gold plating.
Gold plating is the tendency to specify the platforms be equipped every new gadget imaginable, kinda like ordering a new car fully loaded. Take an example from ancient history, the C-5 transport plane. This design goes back to the 1960's. At base, the C-5 was to be the biggest transport that could fly using then new and much more powerful jet engines. It had double the engine power of the contemporary C-141 which yielded a whale of a plane. Flying cargo, including big cargo, to and from real airports (ones with concrete runways) it would have been a very useful machine to have.
It was the first transport aircraft big enough to actually get off the ground carrying a real Army tank, not a light weight "airborne" tank, but a main battle tank, the M1 Abrams. The Army was overjoyed, and began to have visions of future air assaults. The paratroops jump in, capture some bean patch behind enemy lines, then the C-5's swoop in and unload the tanks. Presto, you have a real armored division operating at some strategic spot with complete tactical surprise. Trouble is, landing a whale of an airplane, with a 50 ton tank inside it, on dirt runway doesn't really work. The wheels sink in and the whale is stuck, and blocking the runway. Nothing, not even a tank, is going to tow a stuck C-5 anywhere. It might tear the nose gear off, but it won't move the C-5. Or die hard enemy defenders drop a mortar round on the C-5 as it's unloading and again you have the runway blocked with a blazing and unmovable whale.
To support the "land on dirt runway" requirement, the C-5 was equipped with a fantastic landing gear of 28 wheels. Truly ingenious design solved the problem of retracting this forest of wheels. To go with the 28 wheels, it needed 28 disc brakes, 28 anti skid sensors and an anti skid control box smart enough to figure out which wheel[s] needed less brake pressure and which needed more. I watched a C-5 land at Altus AFB back in the '70's. Touchdown was smooth, followed by flying rubber as some tires blew. At debriefing the aircrew said "Fairly normal landing, we only blew three tires". Antiskid had managed to lock up some of the wheels, and sliding the locked tires along the concrete runway wore right thru the tread.
Then came the requirement for "truck bed loading height", which means having the cargo deck at the height of an ordinary Army truck so cargo pallets could slide off the C-5 onto a truck, no forklift required. Sounds benign. Trouble is, after you get a huge plane that low to the ground, you cannot rotate for takeoff because the tail scrapes on the runway. To solve this problem, the fantastic landing gear was redesigned to "kneel", lowering the entire C-5 to the ground, and then later, jacking the whole plane up high enough to take off.
Then they added a "Maintenance Data Computer" which didn't do much, an unnecessary nose loading door to go with the tail loading door, and a bunch of other stuff that I no longer remember at this remove in time.
Now the C-5 had all the stuff everyone wanted in it. But adding stuff made the plane heavy, heavier than the specification allowed. Lockheed was driven to incredible lengths to trim the weight down to meet spec. This included making the brake rotors (all 28 of them) out of beryllium. Beryllium costs more than gold. When everything else failed, Lockheed made the aluminum skin thinner. This would have fateful consequences later in the C-5's life.
The contract was "firm fixed fee" for a an initial production run (Run A) of 52 C-5's. Lockheed lost barrels of money on each plane it built. When it came time for a contract for the next 52 airplanes (Run B) Lockheed held out for a price that let them earn a little money on each plane. That price was stiff. So stiff that one Air Force general said "I'll haul the troops in gooney birds before I pay that kind of money". His view prevailed, there was no Run B. The 52 C-5's of Run A were all there ever were. They are still flying.
Lessons that should have been learned. Don't specify things that aren't absolutely necessary to accomplish the primary mission. The C-5 was good at moving vast amounts of stuff to and from real airports. Spec'ing in dirt field operation and truck bed loading turned a good airplane into a hangar queen, and sent the cost thru the roof.
These lessons probably have not been learned to this day.
here
He bemoaned the Pentagon procurement system but without offering either an explanation of what is wrong with it, let alone how to fix it. This is unfortunate, so I will attempt to lay out the real procurement problem, gold plating.
Gold plating is the tendency to specify the platforms be equipped every new gadget imaginable, kinda like ordering a new car fully loaded. Take an example from ancient history, the C-5 transport plane. This design goes back to the 1960's. At base, the C-5 was to be the biggest transport that could fly using then new and much more powerful jet engines. It had double the engine power of the contemporary C-141 which yielded a whale of a plane. Flying cargo, including big cargo, to and from real airports (ones with concrete runways) it would have been a very useful machine to have.
It was the first transport aircraft big enough to actually get off the ground carrying a real Army tank, not a light weight "airborne" tank, but a main battle tank, the M1 Abrams. The Army was overjoyed, and began to have visions of future air assaults. The paratroops jump in, capture some bean patch behind enemy lines, then the C-5's swoop in and unload the tanks. Presto, you have a real armored division operating at some strategic spot with complete tactical surprise. Trouble is, landing a whale of an airplane, with a 50 ton tank inside it, on dirt runway doesn't really work. The wheels sink in and the whale is stuck, and blocking the runway. Nothing, not even a tank, is going to tow a stuck C-5 anywhere. It might tear the nose gear off, but it won't move the C-5. Or die hard enemy defenders drop a mortar round on the C-5 as it's unloading and again you have the runway blocked with a blazing and unmovable whale.
To support the "land on dirt runway" requirement, the C-5 was equipped with a fantastic landing gear of 28 wheels. Truly ingenious design solved the problem of retracting this forest of wheels. To go with the 28 wheels, it needed 28 disc brakes, 28 anti skid sensors and an anti skid control box smart enough to figure out which wheel[s] needed less brake pressure and which needed more. I watched a C-5 land at Altus AFB back in the '70's. Touchdown was smooth, followed by flying rubber as some tires blew. At debriefing the aircrew said "Fairly normal landing, we only blew three tires". Antiskid had managed to lock up some of the wheels, and sliding the locked tires along the concrete runway wore right thru the tread.
Then came the requirement for "truck bed loading height", which means having the cargo deck at the height of an ordinary Army truck so cargo pallets could slide off the C-5 onto a truck, no forklift required. Sounds benign. Trouble is, after you get a huge plane that low to the ground, you cannot rotate for takeoff because the tail scrapes on the runway. To solve this problem, the fantastic landing gear was redesigned to "kneel", lowering the entire C-5 to the ground, and then later, jacking the whole plane up high enough to take off.
Then they added a "Maintenance Data Computer" which didn't do much, an unnecessary nose loading door to go with the tail loading door, and a bunch of other stuff that I no longer remember at this remove in time.
Now the C-5 had all the stuff everyone wanted in it. But adding stuff made the plane heavy, heavier than the specification allowed. Lockheed was driven to incredible lengths to trim the weight down to meet spec. This included making the brake rotors (all 28 of them) out of beryllium. Beryllium costs more than gold. When everything else failed, Lockheed made the aluminum skin thinner. This would have fateful consequences later in the C-5's life.
The contract was "firm fixed fee" for a an initial production run (Run A) of 52 C-5's. Lockheed lost barrels of money on each plane it built. When it came time for a contract for the next 52 airplanes (Run B) Lockheed held out for a price that let them earn a little money on each plane. That price was stiff. So stiff that one Air Force general said "I'll haul the troops in gooney birds before I pay that kind of money". His view prevailed, there was no Run B. The 52 C-5's of Run A were all there ever were. They are still flying.
Lessons that should have been learned. Don't specify things that aren't absolutely necessary to accomplish the primary mission. The C-5 was good at moving vast amounts of stuff to and from real airports. Spec'ing in dirt field operation and truck bed loading turned a good airplane into a hangar queen, and sent the cost thru the roof.
These lessons probably have not been learned to this day.
Subscribe to:
Posts (Atom)