Firstly the record's are sanctioned by these people http://www.whpva.org/
The world human powered vehicle association.
There rules are here http://www.whpva.org/WHPVA-CompetitionRules-2009.pdf
Sam Wittingham's records taken from wikipedia are;
Sam Whittingham is a Canadian cyclist who has held several world records on recumbent bicycles.
As of 2009, he holds the following world records under the sanction of the International Human Powered Vehicle Association:
The 200 m flying start (single rider, World Human Powered Speed Challenge, Battle Mountain, NV): 133.284 km/h (82.819 mph) on 2009-09-18.
First unpaced cyclist ever to break the deci-mach mark (1/10 the speed of sound, World Human Powered Speed Challenge, Battle Mountain, NV): 132.50 km/h (82.33 mph) on 2008-09-18. To date, still the only person to have accomplished this milestone.
The 1000 m flying start (single rider): 128.40 km/h (79.79 mph) on 2001-10-06.
The 1 mile flying start (single rider): 126.55 km/h (78.64 mph) on 2001-10-06.
The hour record: 90.724 km (56.373 mi) on 2009-07-17.
In 1993 he also held the record for the 200 m flying start (multiple rider).
These are the record's record's I would imagine that Graeme Obree is interested in..
"The prone position isn't new, and it generally hasn't done well in competition. Aerodynamics depend mainly on fairing design, and the prone position tends to limit chest expansion."
If you look at Graeme's bike design his chest isn't in contact with the bike at any point he is supported solely by his shoulder blades and hips leaving the whole of his chest freedom to move..
Human power even in the best athlete is rarely more than 600w constant
"Direct drive: Using the back wheel itself when it's upto speed to become the biggest chainring in the system ? Has it ever been done before ? Also to drive the back wheel directly via cogs ie NO chain ?"
Cogs are less efficient than chains which have a maximum efficiency of 97%. If you used cogs you would have higher friction losses and therefore less energy transferred to propulsion
"Dual drive cranks: (not sram 9x3 stuff) two cranks left & right on a pista BB (equal spaced) connected to a flipflop rear wheel. What does anyone know about how this idea works and why it never took off ? "
Again less efficient due to the losses of the second chain that's not being driven. Also how can two different sized chain rings linked to one wheel drive at different rates when they are both connected by the crank axle and the rear hub?
"Dual propulsion ie Simultaneous Front & Rear wheel drive, whereby both sets of wheels provide drive/motion to the cycle ?"
Why is dual drive needed when there is not enough power developed to loose traction driving 1 wheel? This would again cause more friction due to dual drive trains.
"Obree NEW bike.jpg"
This again has losses due to having 3 wheels which cause a higher rolling resistance than two.
From wikipedia again;
For example, assuming no wind, one gets the following results for kilocalories required and power delivered to the pedals (watts):
175 W for a 90 kg bike + rider to go 9 m/s (20 mph or 32 km/h) on the flats (76% of effort to overcome aerodynamic drag), or 2.6 m/s (5.8 mph or 9.4 km/h) on a 7% grade (21% of effort to overcome aerodynamic drag).
300 W for a 90 kg bike + rider at 11 m/s (25 mph or 40 km/h) on the flats (83% of effort to overcome aerodynamic drag) or 4.3 m/s (9.5 mph or 15 km/h) on a 7% grade (42% of effort to overcome aerodynamic drag).
Amateur bicycle racers can typically produce 3 watts/kg for more than an hour (e.g., around 210 watts for a 70 kg rider), with top amateurs producing 5 W/kg and elite athletes achieving 6 W/kg for similar lengths of time.
So given that the load i.e. rider and cycle are a constant weight, power is limited to the riders ability and that friction is a constant due to to the tyres rolling resistance and drive train losses the best gains in speed/kg/w can only be achieved by decreasing the aerodynamic drag.
The most efficient aerodynamic shape to reduce drag is the streamlined cigar shape with the widest end at the front tapering to the rear which has a drag coefficient of 0.04 (which is the design Graeme is using) unlike most recumbent's which are generally the same shape reversed which has a higher drag coefficient.
I'm really enjoying this thread and look forward to the reply s its good to get the brain working......