Jump to the beginning

10/03/05- This past weekend we were finally able to film the Formula Turbokart vs. Ferrari video. In the video, we pitted a Ferrari 360 Modena against our Formula Turbokart in a drag race. Follow this link to go to the Formula Turbokart vs. Ferrari video page.

The filming of the video was great fun, but it was especially nice to run the kart hard for hours on end with the new Mark II engine. We loved the way the modified engine idles down nicely at 52%, providing minimal residual torque and allowing the kart to be easily controlled at low speeds. The auto-relight flame-out protection was also incredibly effective and the kart is flame-out proof. And the air start system is terrific; the engine lights every time, starts cooler, and two tanks of compressed air provide enough starts for a full day of fun. If having to choose between external batteries or external compressed air, I'd go for compressed air every time.

Of course, the performance of the kart is breathtaking, or maybe frightening is more appropriate. With the lower engine idle speed, it takes slightly longer for the engine to respond and deliver full torque, but it is still extremely responsive, especially when compared to some of the larger turbine engines. And the power and torque is obscene. If you dare to push the throttle pedal all the way down you are met with an unmitigated rush of thrust that is not comparable to anything else I've ever felt in my life.

At this point, there isn't much else we can do to improve Formula Turbokart, and the filming of the video seems a good place to close the project and consider it finished. We will still drive the kart when we get the opportunity, and there will always be minor maintenance issues, but we are happy to say that, after more than 3 years, Formula Turbokart is completed and is a total success.

...Just in time for my next kart project...Stay tuned for more!

8/30/05- After much delay, we have completed the engine swap on Formula Turbokart, replacing the original engine with our modified, Mark II JFS-100, complete with low idle speed, flameout protection, automatic relight, and air impingement starting capability. We have also put in a couple of quick test runs, and the new set up runs like a dream. The kart is so much easier to control now that the engine isn't constantly fighting the brakes, and my panic throttle chops don't flame the engine out anymore.

Realizing how frustrating it must have been to see a picture of Beavis and Butthead on this page, I have pledged to offer, in the very near future, a video of the much anticipated and long promised Turbokart vs. Ferrari drag race. The most difficult part in making the video will be finding a venue where we can get a couple of renegade runs in without attracting the attention of the local authorities. In case you haven't noticed, Ferraris and jet go karts both share a similar characteristic in that they attract a ton of attention wherever they go. In this case, we want to avoid attention. Nevertheless, I vow to put my neck and my clean criminal record on the line to provide a very small group of web surfers some mild entertainment. Who said I was crazy? (next entry)

9/06/04- The second objective in our JFS-100 gas producer test program was to completely eliminate the flameouts that occurred when the throttle pedal was released quickly. To accomplish this, we adopted a two-pronged approach.

Symbolism, perhaps?

The flameouts were being caused by the engine speed governor closing the metering ports too much, interrupting fuel flow to the burner and causing the flame to be extinguished. We decided we needed to find a way to ensure that a small, uninterrupted flow of fuel would always be available for the burner. We accomplished this by installing a fuel bypass line, completely bypassing the fuel governor, to supply a very small flow of fuel to the burner. We tapped fuel off the engine driven fuel pump via a fuel pressure instrumentation port that is normally not used. We ran the fuel through a .015 inch restrictor valve which we had to drill out slightly to get the required pilot fuel flow, and then rejoined the main fuel flow path at the tee just before the fuel cutoff solenoid, ensuring that there is always a small pilot flame in the burner even when the engine is decelerating quickly. This modification alone eliminated deceleration blow outs about 85% of the time in the test rig.

Line coming off of fuel pump and .015" restrictor valve

To get that number to 100%, we incorporated an auto-relight function into our engine. By wiring the igniter box through the engine's 50% cutout switch and keeping the circuit energized, we were able to have the igniter automatically come back on whenever the engine rpm dropped below 50%, which is just below our idle speed. The igniter will then automatically relight the flame which would bring the rpm back up to idle speed. This modification worked every time in the test rig, and with the two mods together, we have a completely flameout-proof engine, and testing supported this completely. No matter how hard we chopped the throttle, the engine would not flame out.

With these modifications complete, we are ready to reinstall the power turbine assembly, perform one final engine test to make sure everything is working right, and install our new JFS-100 Mark II Turboshaft into the Formula Turbokart. (next entry)

8/22/04- The air start test run is complete and it has been a total success! Check out the videos section for a quick clip of Vito and I starting up the JFS-100 gas producer on compressed air. In the video we are using a manual air valve to initiate air flow to the engine, but once mounted on the go kart, the engine will use a electrically actuated solenoid valve.

Some observations on the air start: Starting with an tank pressure of 3,000 psi, the engine seemed to start more quickly and more easily than with the electric starter. In fact, we did not have one failed start attempt. We started the engine approximately ten times and it always lit off immediately and positively. The cool air blowing across the high pressure turbine blades should also help to keep the turbine inlet temps down during the starts. We were somewhat disappointed by the air consumption of the starts. Each start consumed approximately 350 psi of tank pressure, and when the pressure dropped below 1500 psi, the starts would take longer and would even start to hang a little, indicating that we probably shouldn't try to start the engine with any less than 1500 psi. This was good for about 5 or 6 starts per bottle, which is about half of what I expected. Nevertheless, we used to get similar energy consumption using the 24V battery with the electric starter; after about 5 or 6 consecutive starts with the batteries and starter, the batteries would be pretty drained. As a plus, the tank is rated for 3500 psi, and we will probably save a little air when we use the solenoid air valve in conjunction with the engine's 50% cutout switch, so the actual number in practice might be closer to ten starts per bottle, which would be perfectly ok.

Once we started the engine, we idled it down to 52%, and adjusted the idle stop to maintain that rpm. The pictures below show the engine idling comfortably at that speed. We can tell by the low intensity of the exhaust gas flow and temperature that the engine is producing very little power at this true idle speed, much less than the old idle speed of 68%, back with our old setup. We noticed a couple of hot spots on the power turbine nozzle, but they weren't any brighter than when the engine is at high power, and shouldn't be a cause for concern. At high power the nozzle glows a bright orange.

We ran the engine up to about 90%, but because of a balky tach indicator and some complaining neighbors, we didn't have the opportunity to set the full power stop at 104%. There is still some more testing to do anyway, because we are trying to modify the throttle lever to minimize the possibility of engine flameouts when the throttle is released quickly. The go kart is so frighteningly fast that sometimes it is hard to not snap your foot of the accelerator pedal when you want to slow down. This usually causes the engine to flame out, so we are working on a solution to minimize this problem.

We are very excited by this successful test as we continue our program to make the go kart more driveable around a race track, where we will be doing a lot of testing next season. Stay tuned for more information. (next entry)

8/15/04- Shown in the picture below is the JFS-100 gas producer in the air start test rig. We have already spooled the engine up a bunch of times with the compressed air from the SCUBA cylinder, but we haven't fired it off yet. We will perform a test start and a throttle test in the following week. It will be then when I will set the engine idle and the governed maximum speed. We've had the engine idling down at 52%, so as long as we can get the engine started on air, then our test will be successful and our modification will be complete. Then we will replace the power turbine and swap this engine for the one that is on that go kart that we've been using, so we can perform the mod on that one also. We are running a tank pressure of just approximately 3,250 psi. Check back in a few days for the test results and a video of the test firing. (next entry)

7/15/04- Having fully committed to the air start on Formula Turbokart, we went ahead and removed the electric starter and the AC generator, sealing up the openings left in the accessory gearbox with a plate and gasket. We picked up two pressed steel SCUBA cylinders. We decided to go with steel for extra safety. A typical SCUBA cylinder is an aluminum, 80 cu/ft cylinder. This tank can hold nearly 2/3 more air volume than an aluminum 80 and can do so at higher pressures. Yet, it remains light and easy to transport. We should be able to get quite a few engine starts from each cylinder; I'd estimate at least 12-15 starts each, which should be more than enough for a typical day on the track.

The pictures below show some of the rigging for our connection from the SCUBA cylinder to the kart. From the cylinder, the yoke takes the air out at a maximum of 3,500 psi and runs it out through a long hose and into two ball valves. The first ball valve will be a manual air cutoff while the second ball valve will be an air bleed to atmosphere, which we will use to bleed the residual high pressure air from the lines so we can disconnect our quick connect fitting after the start is performed. From the quick connect fitting, a kart mounted high pressure hose will bring the air to the solenoid actuated air valve, which will open when we hit the start button. Once the start procedure is completed, the 50% cutout switch will interrupt current to the solenoid, causing the air valve to close. From the air solenoid, air will run directly into the engine's impingement duct fitting, where it passes through a check valve and through the impingement nozzles, as shown in the picture on the right.

The pictures below show the Max-Air 35 compressor that we are using to charge our SCUBA cylinders up to 3500 psi. It is not your typical shop compressor. Read all about it on our Cool Product of the Month page. The compressor can charge an 80 cu/ft cylinder from empty to fully charged in about a half hour. It will take approximately 50 minutes to charge one of our larger cylinders. We connect to our cylinder using a standard yoke connection which mates to the yoke fitting on the cylinder.

We expect to have the modification complete by the end of the week. Then we will rig up the JFS gas producer in the test rig, hook up our control panel, and do a test firing. We will set the idle to 52% and then run the engine up to 104%. This should be completed in the next couple of weeks.

Incidentally, just for the heck of it, we hooked up a regular 90 psi shop air line to the engine's impingement connection and tried to spool the engine, and it didn't rotate at all. It is no wonder that such high pressures are required to spin the engine up to lightoff speed and beyond. Stay tuned for more updates! (next entry)

6/28/04- Testing Update: The first four or five test runs of our starter clutch modification worked very well. We started and shut the engine down numerous times in the last few days, and everything seemed ok. We were able to set the idle as low as 49%, though we thought it better to raise it slightly to 52%, without observing any problems with high temperature or flames on the combustor walls and turbine nozzle, except on startup, which is normal.

That's when trouble struck...Feeling more confident about the modification, we decided to start throttling up the engine. The engine ran well up 60%, 70%, and even 80%. At that point, we decided to run it up at full speed. At full speed, even the gas producer without a jet pipe is creating a significant enough amount of thrust to slide the engine stand across the ground, not to mention tremendous quantities of hot gas and uncomfortably loud noise. We spooled the engine down to idle, then shut it down. When we went to start up the engine again, the starter would not respond.

Later tests indicated that the starter was no longer drawing current, which seemed to indicate some serious mechanical damage. We pulled the accessory gearbox off the engine, removed the starter, and found our fears confirmed. The most popular theory is that the sprag clutch wasn't fully disengaging the starter from the engine, and when we spooled the engine up to 100% rpm, it must have been dragging the starter along with it at twice the speed that the starter is rated to run, which basically blew it apart inside.

Woops...

We decided that, in light of this failure after the first run up to full power, that the sprag clutch modification is not worth pursuing. I don't have the time to try to develop this modification or the stomach to blow up more starters. But that doesn't mean that we've given up. I am working on converting the JFS-100 to an air start arrangement. The JFS-100 has an air impingement duct on the side of the engine. Connecting an air line to the fitting allows you to inject very high pressure air through a check valve and to four tiny nozzles which are directed onto the high pressure turbine, causing it to spool up. Using an air start will allow me to remove the electric starter completely, totally eliminating the starter clutch issues, while also saving weight. The biggest challenge will be getting the 2,000-3000 psi line pressure required to start the engine. I decided to go with a SCUBA tank, as they can safely hold 3,500 psi of air in a lightweight and easy to transport package. I picked up a large, 130 cu ft cylinder and an electrically actuated air valve from www.avonaero.com

I will rig up my air system similarly to what is shown in the picture on the right. (Sorry for the lame MS Paint graphics) In this way I will still be able to start the engine completely from the driver's seat, and then simply disconnect from the quick connect fitting and be on my way. I can also rig the air solenoid through the 50% cutout switch so the air automatically cuts out when the engine reaches idle. It should be pretty slick when it is done, and there is little doubt that it will work. (next entry)

6/20/04- After we modified the fuel governor by machining down the fuel governor spring tension adjustment spring plunger, we disconnected the AC generator since we will not be needing it anymore to run the engine. Then we reassembled the engine gas producer section, replaced the stock AC igniter box with a DC exciter, fabricated a simple engine mount, and wired up a basic control panel so that we can test our starter clutch modification.

We plan on executing our first test of the new mod in the next day or so after work. However, I have confirmed that the new all-DC control system works and the sprag clutch does transmit starting torque, as I did a quick spool up and light off in the house. As much as I like the smell of burning kerosene, I don't recommend starting a gas turbine inside a house. Starting the engine manually with the DC control system is more satisfying than just hitting a button and having the engine start up automatically. This way you feel more like part of the system.

If this mod proves itself in the test stand, we will be able to install this engine into Formula Turbokart and we will be able to idle the engine lower which will improve the drivability of the kart tremendously. So far everything seems to be working well. I will report our test results in the next few days.

In the meantime, we have been driving the kart here and there, getting a few renegade runs in before the authorities censure us. The kart is running good but will really benefit from this mod. (next entry)

5/18/04- We came up with a way to couple the sprag clutch to the starter shaft and starter clutch carrier in a way that I feel will effectively transmit the starting torque. We started by pulling off the over-running clutch carrier, and reinforcing the carrier with a machined steel ring which we pressed in place. Then we machined the inside diameter of the clutch carrier to accept the outer race of the sprag clutch. Then, we removed the splined ratchet wheel from the electric starter output shaft, machined the teeth off of it, and then machined it down so it would fit inside the inner race of the sprag clutch.

The inner and outer races were then applied to their respective carriers using a very tight press fit and a gear and shaft retaining adhesive compound. I am currently in the process of re-assembling the gearbox. I plan to re-assemble the engine gas producer, wire up a basic control panel, and perform about 25 test engine starts before pulling the engine apart and inspecting the clutch assembly for any signs of problems. Only then will I consider the operation successful. (next entry)

To see the previous part of this project, click here