After purchasing the kart it needed a little cleaning up and some work on the front brake master cylinder ,OE seals were going to cost ~$70 for a set , the local brake repairer sold me 2 sets of 7/8" dia seals for $17
A standard karting engine mount platform was purchased to mount the freepower on so that things were as "original" as possible with regards chain alignments etc , the original engine was a 125cc with 6 speed gearbox running heavier duty chain to the axle so new sprocket carrier and sprocket (81 tooth) were needed for the axle , some extension of the axle keyway was also required so as to bring the sprocket into alignment with the freepowers sprocket .
The actual freepower wheel used is a Cummins "W shaft " salvaged from a turbo repairers scrap bin , they'd thrown it out because of slight impact damage on several inducer tips making unsuitable for fitting back into a 75,000 rpm turbo but quite OK for a 20,000 rpm freepower .
The Cummins wheel ( # 214806 .... I think thats the part number ) turns in the "right " direction for a direct chain drive , if a single stage gear redux were to be used then an opposite rotation direction would be required , a large Garrett turbine shaft/wheel such as #442208-0001 could be used , this would allow even greater mass flow to be used , some ~40% more, allowing a gas producer flowing up to a max of perhaps ~1.5 lbs/sec , perhaps a Cummins ST/VT50 or Garrett TV81 turbo for the gas producer .
The Cummins wheel needed a boss silver soldered onto the compressor quill section , then suitably machined to accept a # 6303 ball race ( no seals) and a tapered section with keyway for the 10 tooth karting engine sprocket , the original Cummins comp nut thread being used for retaining the sprocket on its key.
A 3/4" ID #9R-12 ball race without seals being used at the turbine end of the shaft , both bearings fitted into rebates in the shaft tunnel that was held in alloy clap mounts secured to the original engine mount holes .
Now came the worst part of the entire build , constructing the freepowers scroll , to minimise weight I decided to make a sheetmetal scroll .................bad decision ..... sheet metal warps and buckles every time its sees the welding torch , the scroll was thrown in the rubbish bin several times during its manufacture only to be dragged out a few days latter once I cooled down and realised it had to be finished somehow , with lots of bad language and several attempts at welding , finishing up with some re machining of the exducer shroud area , it was completed .
Scroll A/R ended up being ~1.68 , this is a very large A/R for this sized turbine wheel but because the wheel was being used for a relatively low speed (rpm) application, processing low density gases, it was applicable .
Scroll was fixed to the shaft tunnel using the standard Vband arrangement .
The 2 red plastic covers on brass fittings at the sprocket end of the shaft tunnel are where bleed air from the gasifiers combustor and bleed lube from the main lube pressure system are supplied to the shaft tunnel .
Lubrication is total loss , a supply of a few drops /min is all thats required , the bleed air forces the lube thru the bearings and out into the turbine ducting .
The bleed air also supplied an "air seal" around the shaft behind the sprocket to prevent lube leakage , the majority of the bleed air exiting at the turbine wheel prevents hot gases entering the bearing space , keeping bearing temperatures at reasonable levels.
Total weight of freepower stage including mounts was 8 kgs -~18 lbs
The velocity triangles for the freepower suggested a horsepower of ~32 shp at 15,000 rpm ( ~60 mph ) when flowing ~1.1 lbs/sec - 0.5 kgs/sec at ~650 deg C ITT , with a pressure ratio of ~1.68 :1 across the stage , thats 10 psit as gauge total interstage pressure at sea level .
Good quality racing kart chain can be used on IC kart engines running up to ~22,000 rpm , the "rough" running IC engine will undoubtedly give the chain a more difficult time than a smooth running freepower turbine engine .
If the Cummins wheel in this setup was run up to 22,000 rpm , horsepower would have been well in excess of 40 hp with a speed of >90 mph using the 8.1:1 gearing , slightly lower gearing would produce a potential 100mph kart , maximum speed I was able to achieve on my rough "back road" test track was ~70mph before it became too rough to maintain any sort of safe drive , the kart was still accelerating hard at this point .
Once the freepower stage was constructed such that it provided clearance between it and rear tyre , as well as the driver , it was time to position and mount the gasifier turbine scroll so that the scrolls outlet was nicely inline with the freepower scroll inlet for a "straight in" approach with the hot gases
Due to the weight of the 1.23 A/R turbine scroll and the rest of the rather large Garrett TV84 turbo , the mount was constructed to incorporate the karts twin main axle bearing mounts .
2mm thick steel sheeting was used for the majority of the mount , the sheeting parts were then welded to a thick steel flange with studs that the scroll bolted to , on the other side of this heavy section there was mounted a circular flange with Vband angle ( 20 degree) section on its perifery to which the matching half on the base of the combustor mated into a rebate with locating pin for positioning .
The combustor was constructed from a ~7" dia "2 gallon" air over water fire extinguisher , the thin walled ( <1mm) stainless can is normally pressurised to ~100psi and tested to ~300 psi .
Although the combustor requirements for the TV84 did not require such a long combustor it was decided to leave it as long as possible and use the combustor as a "torque arm" to stiffen up the turbo mount by adding triangulation down to the axle bearing mounts and seat at the far end of the combustor
With the combustor positioned it was time to get an oil tank into the chassis .
Due to the requirement for low weight it was thought that a lube supply with ~5 litres capacity would be they way to go as long as there was adequate heat exchanger cooling (automotive transmission cooler used ) within the system .
The oil tank incorporated an automotive IC engine derived mesh pickup filter of ~4" dia displaced ~half an inch above the tank bottom.
There was also "baffling" within the tank to reduce oil movements during high cornering loads , the low profile of the tank was a problem , ideally a taller deeper tank would be the choice but the low positioning of the TV84 turbo meant a low tank inlet position if gravity drainage was to be the method used .
A deairation plate with multiple ~1/4" dia holes was positioned just below the lube inlet port .
To mount the oil tank a thick alloy sheet metal "skid tray" was fitted at the rear of the chassis
LOL......yeh, karts are a bit of a tight squeeze , I ended up with painful hips for months after the kart testing , my long legs took some folding to get into the kart , getting out was even harder with a hot freepower scroll besides me .
Theres some video of the kart runs on Youtube www.youtube.com/watch?v=F-fci_AlNMU , acceleration was kinda slow initially , as with most single ratio freepowered turbines , but picked up once the freepowers rpm rose and it started producing horsepower rather than just torque .
From a standing start it took ~5.1 seconds to cover the first 50 metres-55 yards , but only 2.4 seconds from 50 to 100 metres .
The kart is now with a new owner , but I'm in the planning stage for another , using either my 9/94 engine, www.youtube.com/watch?v=8Y8jpaDTkEY , with afterburner, or a "fan " engine I'm considering building using a transonic first stage axial comp wheel flowing ~10 lbs/sec powered by a 4th stage Allison C20 freepower wheel with gas energy coming from the 9/94 engine or a DIY gas turbine engine made from a new/reman Garrett GT60 turbo I have sitting in the shed , the GT60 turbo weighs a lot and would probably require something more substantial than the usual lightweight kart chassis .................LOL, mind you , both setups will need something substantial for the projected ~500 lbs of thrust, project for 2011 year :-)