Holset HX50 Supermoto Jetbike

[slittlewing]

Hi there!

I'm Scott, and have been pointed to this site by Andy after he saw the video of my DIY Gas Turbine Engine on Youtube. He has pointed me towards this forum so I thought I would do a "build thread" although the jet is already built (better late than never). 

I haven't had any help thus far since building the engine so any advice or comments on things that could be improved would be most useful! I have not done any calculations in building it, just sized things on what "looks right".

The build started when I obtained the "sledge" rails from work after they were being chucked out. I thought I could make a rather attractive sledge and then thought.. why not build a jet engine for it! Always fancied making/owning one as they are quite a spectacle.



Heres the details:

TURBO
The turbo is a Holset WH2D model, I think its the largest they make and originally from a 10L Cummins Diesel. I don't know much about its history but it had a rebuild tag on it when purchased and there is a normal amount of bearing play.

The compressor inlet is 4" and it weighs 25kg, I think it can deliver 3 bar boost in "normal" use but I cant find a compressor map for it.





COMBUSTION CHAMBER
Whilst everyone seems to build their combustion chambers in direct line to the turbine flange (which I realise is most efficient to not lose energy from the expanded/hot gas), I wanted the combustion chamber to be invisible to the eye on the sledge for aesthetics, not poking up vertically. I therefore "underslung" it, forward of the turbo. The the hot gas has to make a 90 degree turn to enter the turbine flange in this case.






The combustion chamber is made from 5.5" steel pipe, I went for a 5mm wall thickness. I know this is massive overkill, but I wanted complete confidence it wouldn't blow up when weakened/heated! I had the turbo flange and end flanges laser cut from CAD which I did.



The inlet to the combustion chamber is 2.5" steel pipe and connects to the turbo via a "V band" clamp. I angled the air inlet to the combustion chamber to enter at a tangent and create "spin" of the air around the flame tube.




Being quite narrow at 5.5" (to fit under the sledge with some ground clearance), I made it quite long at around 800mm. I'm sure it could be half the length but I thought there wasnt much disadvantage other than weight, it just allows for more time for the flame to burn.

FLAME TUBE
Probably one of the biggest unknowns was the flame tube. I know that there is some science if you do it properly, using a tube with certain amounts of holes and spacing. I decided to take the easy option and use perforated sheet rolled into a tube, purely because I had seen a website where it had worked for someone else. My thought was that whilst there is a much greater "hole" area than required, the flame can draw in as much air as it wants to.

I also added some "swirl vanes" to the outside of the flame tube (laser cut). This was firstly to try and keep the air spinning around the flame tube (in my head, pinning it to the outside via centrifugal force unless drawn by the flame), and secondly it provides some blockage to the airflow. I thought this was beneficial to create something for the turbo to push against (as in a vehicle engine) and not create something too "free flowing" which would mean the turbo would end up overspeeding to actually push enough airflow to make boost.




I also blocked the first few inches of the flame tube (where the compressed air enters) to avoid the fresh air blowing across the fuel injectors.

SPARK PLUG/IGNITION
I positioned the spark plug at the very back of the combustion chamber, before the turbo. I wasn't sure if it was too far back (and the flame might just blow out of the turbine) but it does work and allow the flame to propagate back into the flame tube.
 


I use a relay with a capacitor to generate a pulsed voltage, which goes to a Bosch Ignition coil driver (from a Mk2/Mk3 Golf etc), which goes to a Bosch coil.



FUEL SYSTEM
There is a propane injector, with a quick disconnect pipe at the front of the sledge. I have a propane bottle with an adjustable 0-4bar regulator for starting the engine along with a leafblower.

There is also a wet fuel system (petrol). I use a Marelli car injection pump (3-5bar) with PWM control to alter pump speed and therefore fuel flow. There is no pressure regulator in the system. For the injector, i first tried using a 0.8mm carb jet. When i tested it into a bucket, it created a single "jet" of fuel like a super soaker, not a spray. I therefore adapted a car mechanical fuel injector (from a Bosch CIS/K-jet system) instead. The bad thing is that it doesnt open until about 4bar pressure, at which point the pump is driving quite hard against zero flow. 



There is a high pressure fuel filter after the pump, and a gauze filter before it.

For the afterburner, at the moment there is a 12v solenoid which allows fuel flow direct to the tailpipe again through a carb jet. I will be changing this.. see below. There is a cooling loop on the copper pipe between the solenoid and tailpipe, to avoid heating the solenoid.




OIL SYSTEM
The oil system is a Marco 4bar bronze gear oil pump, controlled via PWM to give me pressure control. There is no oil cooler yet, I thought I would do without for short running. I am using Shell "Aeroshell" Turbine oil (for an aircraft jet engine). 
The turbo oil outlet is horizontal, which I know should be vertical. I tried to remove the turbine and clock it round, but it was seized to the bearing housing so I kept it as-is. My thinking is that as long as there is positive oil pressure, oil will flow through the bearings.

The oil tank is about 2 litres and is plastic (could do with an upgrade to an aluminium one).

OTHER
The tailpipe geometry is guessed/based on what cones I could get on ebay. I think it needs a smaller outlet cone.
The battery is a 7AH Lithium Motorcycle one, just big enough for about 20 mins running.



One of the greatest challenges was packaging everything under the sledge... its tight!







I have gauges for boost pressure, fuel pressure and oil pressure, Battery voltage readout and on off switches for the pumps, afterburner and whole 
system. The pod is laser cut and bent/welded.

[slittlewing]

FIRST TESTING
I have so far completed one "test" of the engine. At first, I could not get it to fire up. It would just emit a "bang"/explosion after a removed the leafblower each time. After some thought, this was due to not enough propane and therefore ignition only happening when the airflow reduced and the correct air fuel ratio existed.

I drilled out the propane injector from 0.8mm to 4mm and it fired straight up!

The engine sits happily at 1-1.5 bar boost, but I cant get anymore on propane as the bottle pressure equalises against the boost pressure and the fuel flow drops.
I tried wet fuel, it worked a little at one point but after turning it up it put the flame out. I dont think I have a fine enough mist of fuel, and have therefore ordered some nozzles in different sizes (0.3mm to 1mm hole) which have a pin pointed at them to create a mist. I will update on this when I have got them and tried one in a bucket (they wont arrive for another month).

I intend to use one of these mist nozzles for the afterburner although i guess misting is less critical there.

I would have tried the afterburner on the first fireup but the solenoid has stopped working, so I have a new/different solenoid to fit and hopefully try out.

The aim is to get it running at 3bar on fuel and afterburn on top. I have no idea on the thrust or how I can maximise it, I look forward to the expertise of the forum!

As for the sledge itself.. its entirely unpractical here in the UK (more of a fancy engine stand), unsure whether to mount to a kart chassis so I can test it or to move onto a different project!! (This one has ended up alot more expensive than initially planned!)

Heres the youtube video:



<iframe width="560" height="315" src="https://www.youtube.com/embed/0-b0mJDA5-c" frameborder="0" allowfullscreen></iframe>

[slittlewing]

youtu.be/0-b0mJDA5-c

[racket]

Hi Scott

Welcome to your new "home" :-)

Where can I start ...............what a nice build ..............yep, it would have been expensive .

Theres a number of items that are collectively contributing to your overall performance difficulties.

The first one is the flametube/combustor , it has to be built to fairly conventional proportions/methods otherwise it won't work to any degree, the engine might start and run but thats about all .

Have a read of these rough guidelines first jetandturbineowners.proboards.com/thread/680/diy-turbines ..............then we'll start to go through the various components one by one .

Congratulations on getting it built and fired up, just need to get her "tuned" :-)

Cheers
John

[slittlewing]

Thanks for the response John and the link, lots of good info.

I have taken some measurements, and focusing on the combustor/flame tube first of all, mine is narrower than advised.

Compressor Inducer Dia = 2.5"
Flame Tube Dia = 3.5" (According to the link, should be 5")
Combustor Dia = 5" (According to the link, respecting 1/2" gap should be 6")
Combustor Length = 25" (According to link, should be 15" according to link)

Without building a whole new combustion chamber (which would basically involve re engineering the whole thing), the best I could do would be to create a new flame tube of 4" diameter (still leaving 1/2" clearance to the walls) with the advised hole sizes/spacing?

Cheers,

Scott

[racket]

Hi Scott

With a 2.5" dia inducer of 4.9 sq ins area , if we go for a minimum 3 times inducer area for the flametube cross section then 14.7 sq ins required , so a 4.3 " ID for the flametube , anything less and I can't guarantee you won't have problems when trying to burn kero , you'll get away with burning propane in a skinny flametube though.

Also your elbow at the end of the combustor , it needs to provide a smooth entry into the turb scroll , the hot gases are moving pretty quickly at that point , any "roughness" will produce turbulence and "restrict" flow , also how have you blanked off the wastegate ??

As you've read the RoTs I won't need to go into the fuel delivery/presentation , but with an undersized FT diameter it'll need to be as good as possible.

Its a tad difficult from your pics to see how the FT fits into the outer can , but you need all air going through the FT wall holes , no bypassing of air through "gaps" , also too much swirling of the air going along the FT wall gap is only going to produce a lot of frictional losses as well as making it more difficult for the air to get through the holes, you have a tangential delivery tube entry ( which I don't favour) that will provide more than enough swirl , too much swirl and you can end up with "problems" , we really only need it at the Primary Zone , not further downstream .

Your current flametube probably has 30 times the hole area thats required , its a wonder it worked at all , probably your intermediate swirl vanes along the wall probably created enough restriction to flow that only the top section was actually "working" .

All the best with the mods :-)

Cheers
John

[slittlewing]

Thanks for taking the time to give such an informative reply John!

You have me putting some thought into completely redoing the combustor from scratch haha. Interesting also that the flame tube needs to be closed so all air must pass through the holes, I hadn’t realised that either (mine is flared out at the end to meet the inside diameter of the combustor tube but it’s not air tight)

Please can you confirm whether the ideal FT diameter should be 3 times the inducer diameter, or 3 times the inducer area?
The guide says diameter, which would put me at a 7.5”FT (8.5” combustor) in an “ideal world”?

I have read the injector section and that I should not be injecting the gas axially (which I am) so that would need changing. For the petrol, I would still like to use a single misting nozzle if possible.

[Deleted]

Hi Scott,

PETROL ? i hope you meant to put Kero :-) you will need an oil burner nozzle, type used for big oil heating systems

All best Andy

[racket]

Hi Scott

I normally use 3 times AREA , Jetspecs was setup using diameters because it was easier for guys to understand and for the programming by Jesse..............area is a better method as its more linear with mass flow through an inducer area , whereas diameter isn't .

Your intended 4" FT might be OK if the fuel presentation is "sophisticated" , nice high pressure ( >250psi) with good atomisation, or having vapourisers with lower pressures , you will need to allow for the FT to expand more than the outer can , so some sort of "rattly" slipjoint required.

Please don't use petrol , its not a "good" fuel , plus its not the safest due to all those combustible fumes , use kero or a diesel/petrol mix (80/20)

Yep, we'll slowly work through your design to get it functioning :-)

Cheers
John

[Johansson]

Nice sled!

One thing that concerns me is the oil ports on the turbocharger, it looks like the drain port is pointing to the side of the turbo and not straight down. If it were a ball bearing turbo you might have been ok but a journal bearing turbo needs a gravity drain for the oil so the drain port has to be pointing downwards.

Cheers!
/Anders

[slittlewing]

Many thanks for the input chaps! I had been trying to put petrol in but it sounds like I will need to find some kerosene or try the 80/20 mix! (Newbie error!)

Given my existing combustor would be hard to "seal" with the FT/attach a rattle joint to, It might be easier to do a new one from scratch.
In this case, what would you advise for ideal new dimensions? Would it be best to go for a 5" FT (slightly larger than the 4.3" calculated) inside a 6" combustor?
6*Inducer Dia would put the recommended FT length at 15", or go a little longer?

At the moment I have an EFi fuel pump (3-5 bar), what type of pumps are you using to generate >250psi/17 bar? Or in the case of using evaporative tubes, I presume there is a restricting jet at the top of each tube?

The horizontal drain on the turbo is not ideal as you point out - I tried separating the housing to "clock round" the bearing assy on the bolts but it was seized up! I could re-orient the turbo with a new combustor design though.

Your help is greatly appreciated,

Cheers

Scott

[turboron]

Scott, after taking part several turbochargers I found the best method for "frozen turbines". Step 1, heat the turbine housing with a propane torch as hot as you can get it within 10 minutes or 15 minutes then let it cool to room temperature then repeat two more times for a total of three cycles. Step 2, on the small turbines I use the turbine retention bolts can be backed out against the oil drain etc and used as a jack to load the turbine during the heating cycles. Step 3, suspend the turbo (with a short piece of chain rapped around the compressor housing on the bearing housing side) from a chain hoist or other means. Heat the turbine housing again with the propane touch and strike the turbine housing near the flange with a punch and a heavy hammer. I have used this method several times with complete success.

Other members may have better techniques.

Thanks, Ron

[racket]

Hi Scott

If you can squeeze a 6" OD outer can into your chassis that would be ideal , but no need to go bigger than 4.3"ID for the flametube , by the time we take wall thicknesses into account you'll have a nice gap between them .

The problem with EFI pumps is their lack of pressure , if theres 2 bar of air pressure within the flametube then even with 5 bar of fuel pressure theres only a 3 bar pressure drop across a spray nozzle , which isn't going to provide anything other than a fairly large atomisation , EFI pumps are better suited to vapourisers that feed metering jets ( syringe needles) at each vapouriser inlet .

Some of the high performance EFI pumps give more pressure, but one really needs a small "hydraulic" gearpump for high pressures ( 500-750 psi) and fine atomisation .

Overall length of the new FT can be 15", though thats the "working length", so add on your blanked off length at the top where your delivery tube is fitted.

Cheers
John

[slittlewing]

Thanks for the feedback - Ron I will have to try your heating/cooling cycling trick again - last time I don't think I struggled to get the turbine housing hot enough with my piddly blowtorch.

I have modelled a new flat top/bolted flange Combustor with 6" outer and 5" FT, proportioned the holes to the inducer volume (30/20/50 as mentioned elsewhere on the site). Do you think this design would be optimal for Kerosene?

I would be missing the "domed cap" and swirl vanes many people integrate into their designs - not sure how much of an issue/performance loss that would be? I could try modelling a domed cap FT design but it does complicate the fabrication a fair bit.






On the fuelling, I think I will try to use an 044 Bosch Pump with kerosene (aka copy andym)!

I look forward to your comments,

Cheers

Scott

[racket]

Hi Scott

A couple of points , if your outer can is 6" OD then ID will a tad smaller , and with a 5" OD FT your gap might be getting a tad small , especially where the delivery tube meets the can , trying to get the air out of the delivery tube and into the gap around a right angle corner will prove difficult , I'd be dropping the FT OD back to ~4.5" max , 4.3" ID to provide a bit more clearance .

Also , it might pay to add some holes at the very top of the flametube to blow air across the underside of the flat "lid" which is exposed to high radiant heat and can start to glow red .

Cheers
John