Kamps Micro-Turbine Build

[Munkul]

It sounds pretty straightforward as I write it here, but there was a heck of a lot more faffing than the description suggests!

succinctly put

I like your bearing solution very much. I wondered before I read the rest of the post - whether there was some special high speed cage option.

Out of interest, why not go down the route of making a brass journal setup and using an oil pump for pressure lube? It seems to be the simplest setup for turbos anyway.

 

[Misterg]

Out of interest, why not go down the route of making a brass journal setup and using an oil pump for pressure lube? It seems to be the simplest setup for turbos anyway.

Thanks

Basically, I'm just following the design in the book (with the odd tweak) - I think the main intention for the design is an engine that can be used to power an RC plane (though I'm not intending to do that). I guess things like a separate oil pump and tank would just add weight, bulk & complication. (If modifying a turbo, I think that would be 100% the way to go though.)

Current practice for this sort of engine seems to be to bleed air from the main casing into the shaft tunnel and use this to carry a kerosene / oil mist through the ball bearings to cool and lubricate them. The same mixture can be burned as the main fuel (like premixed 2 stroke fuel) and pressure can be taken from the main fuel supply so no separate oil tank or pump is needed (just a small bleed off the main fuel supply). The oil / kerosene coming through the shaft tunnel just gets burned up in the exhaust. I'm sure it isn't good for fuel efficiency, but these are hideously inefficient anyway!

Glad people are interested.

 

[optima21]

Ive got to say that Im slightly disappointed by some of the contributors on this thread

as there was copy of the book on ebay, I was expecting to be in a bidding war for it

but no-one bid against me so it was mine for £1.00 (plus postage)

 

[mike os]

Ive got to say that Im slightly disappointed by some of the contributors on this thread

as there was copy of the book on ebay, I was expecting to be in a bidding war for it

but no-one bid against me so it was mine for £1.00 (plus postage)

Dammit I paid a tenner on amazon

 

[Misterg]

After finishing the shaft, I mounted the turbine, tightened the nut up, and put it back between centres to skim the blades to diameter. I thought I should quickly check the runout before starting to machine and it was miles out! (about 0.05mm). I slackened the turbine nut and it went back to ~0.004mm. I dismantled it again, gently stoned all the mating faces, gave it a good clean in fresh solvent and it went back together with~0.004mm runout still - looks like one needs to be *very* careful with these things!

Next the shaft tunnel and diffuser

These are made from 6082 aluminium. Both are static parts and form the 'backbone' of the engine - the shaft tunnel holds the shaft bearings and mounts to the centre of the "diffuser" (This is a series of vanes that form air channels that gradually increase in area. They take the air that is spun off the edge of the compressor wheel at high speed and gradually slow it down, which increases its pressure, and in the process, turns it through 90 degrees towards the combustion chamber.) In this design, the vanes are made separately and fitted into slots in the main "hub".

The shaft tunnel needs bearing seats at each end that are true to each other, and to the mounting flange.

I set the stock up with tailstock support and machined the (parallel) rear section first, taking most of the stock down to the finished diameter of the parallel section:



I replaced the tailstock support with a fixed steady running on a part of the tunnel that would later be machined away and machined the rear bearing seat and drilled the central hole part way through.



This gives a part where the rear bearing seat is concentric and true to the OD and the rear face.

Flip it around in the chuck and grip it carefully on the finished rear section. Set the OD running concentric and true and the rear bearing seat is automatically true.

I needed the steady again to machine the front face, front bearing seat, the locating diameter and to finish drilling / boring the through hole.

I could also drill the tapping size holes for the three mounting screws using my toolpost spindle.



I'd already made a plug to fit in the front bearing seat that would take the tailstock centre to hold the part while cleaning up the outside of the tunnel (it's not essential that this section of the OD is absolutely true to anything).





There's a non-critical curve at the front of the section. I used this as a test piece for machining a curve to co-ordinates ('machining by numbers') as there are a few critical curves that will need machining later on. It proved my spreadsheet worked, and that I could use a 3mm radius tool without too much chatter.



I machined the rear of the diffuser "hub" first. This was trickier than I though it should be due to the need to switch between tools to get into the various contours.

The three radial grooves form the air bleed into the shaft tunnel to cool / lubricate the bearings. There will be a small pipe trapped in one of the grooves that dispenses an oil/kerosene mix into the airflow. The cooling air grooves were milled using my toolpost spindle while the part was still mounted in the lathe (along with drilling various holes - it really is handy!).



The part was mounted on a mandrel to machine the front face to try and keep the front and rear faces parallel. I used the 3 mounting holes initially, as I wanted to keep access to the centre point for marking out the diffuser vanes after the outer part had been machined. (I was pleased to see that there was almost zero axial runout after flipping it around.)

I thought I'd be clever and use a drill to gently mark the centre of each vane to help with marking them out - the centre of the radius falls just inside the edge of the diffuser disc. I though I would lose the marks when machining the radius on the outer edge of the diffuser. I would have done if I hadn't drilled too deep though . I suspect the marks are annoying, rather than a problem, but I can fair them with epoxy later.

While I was at it, I tried to be really clever and drill a 1mm hole where each diffuser vane would met the edge of the 'hub' in the hope that this would give me 'half a hole' to give the saw somewhere to start when cutting the slots for the vanes. In another brain fart, I machined these *inside* rather than *outside* my reference diameter. (Luckily they are still on the line of the vanes.)

To add insult to injury, I managed to break a 1mm drill deep in one of the holes.. Hey ho...







The outer edge of the diffuser is tapered with a 3mm radius on the front edge. I cut the radius using a router cutter as a form tool which worked surprisingly well.



I then drilled and tapped the centre of the mandrel so I could clamp the centre of the hub to machine where the three mounting screws were and countersink their holes. It was then re-mounted it using countersunk screws to take a light skim off the area behind the compressor.






The rear bearing seat has a recess for the preload spring. I have got some wave washers for bearing preload - they have quite a high spring rate (~30N/mm) so I have used two (separated by a spacer) to try and make the preload less sensitive to small changes in dimensions - 1.6mm installed height each should give me the 15N recommended in the book, so an appropriate spacer was made



I could then have a trial assembly





The front spacer needs adjusting to bring the edge of the compressor wheel level with the face of the diffuser. Feeler gauges and a fingertip seemed to be the best way of measuring the height of the compressor to see how much needed to be machined off the front spacer:



With a new front spacer...



...the edge of the compressor disc is, as near as I can tell, level with the surface of the diffuser.





(The shallow holes in the edge are where I drilled too deep when marking the centre of curvature of the vanes; If the deep holes had been drilled right, they should have been split by the edge of the hub, but they'll be cut through anyway when I make the slots for the vanes.)


Thanks for reading

(Oh if anyone is buying the book, there are a couple of errors to look out for...)

 

[Maker]

How long did it take to interpolate that radius? I'll have to try that, usually just bodge curves with form tools and files.

 

[Misterg]

How long did it take to interpolate that radius?

I can't remember specifically, but I would guess ~30-40 mins from start to finish.

 

[Munkul]

Got any more details of that toolpost milling spindle? Sounds like a really handy tool!

 

[Misterg]

Got any more details of that toolpost milling spindle?

I’ll stick something up on another thread

 

[Misterg]

I cut the curved slots for the diffuser vane with a simple jig on my hobby woodworking bandsaw. It's definitely not designed for this, but it worked a treat and produced a kerf that was a lovely fit on the 1mm sheet aluminium vanes:



(The pen is pointing to the metal pin that the top part of the jig pivots on.)



I did find the remnants of my broken 1mm drill bit, which cost me a bandsaw blade.

I thought there might have been a piece of it left in the hole, but couldn't see it or remove it (the drill bit shattered, rather than just breaking in two). Having cut into the hole (and wrecked the blade) I could now see there was a fragment there.



A few hours of boiling the diffuser hub in aluminium sulphate and a lot of picking and fiddling later, I eventually extracted the culprit. This is it sitting on a razor blade (it's 1mm diameter, remember? - little sod!):



I used the bandsaw again to rough out the vanes, then stacked them up and milled them to size.







To form the curve in them as evenly possible, I used a hose clamp (and a cordless drill) to pull each vane down onto a former - a 4" diameter tube gave the 80mm radius required for the vanes after they sprang back.







I was sorely tempted to try tacking the vanes in from the rear with the TIG welder, but got worried about the possibility of distorting the 'hub' so chickened out and glued them in, as per the book. The problem here is that the diffuser can get to over 100°C, just from the heat generated by compressing the air, so the glue needs to be able to withstand this (the book just says something like 'use high temperature epoxy'). The front cover clamps down onto the vanes and people have reported the the vanes pushing through the diffuser 'hub' due to the glue (possibly standard epoxy) softening. To help avoid this, I had made the vane slots slightly shorter than required, and machined an undercut at the leading edge of the vanes so that this part of the vane gets sandwiched between the diffuser 'hub' and the front cover to take the clamping pressure (rather than relying completely on the glue).

I spent a while trying to find a temperature resistant epoxy that didn't cost an arm and a leg. I eventually went with UHU Plus Endfest 300 epoxy (no I hadn't heard of it either!) The data sheet gives information on mixing ratios and curing temperature to maximise temperature resistance and gives strength data to 140°C.

Before fitting the vanes, I cleaned out, filled in & faired the redundant holes in the diffuser 'hub' using metal loaded epoxy. The vanes themselves were glued in using the Endfest epoxy, cured at ~120°C.



(I see that burr - it will be removed!)





The vanes need cutting down to an accurate height above the diffuser body (to match the compressor wheel tip height); and the outside profile needs to be cut to match the inside of the casing / front cover.

I potted the vanes in wax to stabilise them during machining (mostly paraffin wax, with a bit of beeswax to make it sticky).



I'd left the mandrel in the lathe chuck from when I turned the 'hub' so getting the assembly running true was straightforward. After getting them to the correct height / overall diameter, I used more 'turning by numbers' to get the 8mm radius needed on the outside edge of the vanes (where the air turns the corner).



I think the wax shows the way that the size of the gap between the vanes gradually increases really well.

Most of the wax got melted off with a hot air gun, and a scrub in hot water and Fairy liquid got rid of the residue.



I needed to spot face the back of the diffuser where the front cover securing screws sit, but otherwise, I'm leave the rear of the diffuser as it is - there's about 0.5mm of vane proud of the surface with a good fillet of epoxy on it - the rear corners of the vanes are cut away so that there is a smooth transition at the edge of the diffuser.

I also thinned down the trailing edge of the diffuser vanes, but I'm not sure how worthwhile that is...



The front cover securing studs pass through 3 of the vanes so I needed to carefully mill away the vanes to match the holes already drilled in the 'hub' (I'd partially cut the vane before gluing, but left enough metal to keep it in one piece whilst being glued up).



The securing studs are supposed to be thinned down and faired into the vanes, but I'll leave that for later!

Back to sheet metal work next, I promise.

Thanks for reading!

 

[Dr.Al]

I'm in awe of this build. I think I learn some new technique with every one of your posts.

 

[optima21]

I'm in awe of this build. I think I learn some new technique with every one of your posts.

yeah just what I was thinking too, its not what he's making, its the way he's making it, and the thought process involved with adapting the equipment available to make the parts he wants. I think most of us would have given up before we even started cutting up pieces of metal.

 

[Prajna]

Damn, I'm not seeing the images from the last couple of posts, MrG. Usually I don't see them the day you post but they turn up the following day.

 

[kappi]

I just discovered this forum. I am enthusiastic about Misterg's skills and the way he does his work. 25 years ago I also built a turbine according to the plans of Thomas Kamps and got it to work more or less. In the meantime I have built 3 KJ 66 turbines and fly them in model airplanes. At the moment I'm also building a Kamps turbine again, mainly because I still had a few parts left. Here are a few pictures, but of course not comparable to Misterg's quality ... Greetings from Germany
Kappi

 

[Misterg]

Thank you!

adapting the equipment available to make the parts he wants.

Having seen some of your posts, you should know!

The challenge / enjoyment for me is making what I want with what I've got. I'm making most of this up as I go along!

@Prajna - sorry the images aren't working. I've sent you a PM as I would like to sort this out.

 

[Misterg]

25 years ago I also built a turbine according to the plans of Thomas Kamps and got it to work more or less. In the meantime I have built 3 KJ 66 turbines and fly them in model airplanes. At the moment I'm also building a Kamps turbine again, mainly because I still had a few parts left. Here are a few pictures, but of course not comparable to Misterg's quality ... Greetings from Germany
Kappi

Hi Kappi and welcome! Thank you for posting - your work is obviously higher quality than mine, as your engines have actually been proven to work

I get the impression that there is more interest in these engines in Germany / NL than the UK.

I think I would have gone for a KJ66 if I could have found a cast turbine disc at a cheap price, however given the amount of work that has gone into the engine so far (it has surprised me - it takes a lot longer than I was expecting) maybe a Chinese made turbine disc would be worth putting on the shopping list. I will see if I can get it to run with my homemade one first.

I don't fly, but have seen videos of RC jets that are pretty impressive. They must be awesome in real life.

 

[kappi]

I have no doubt your turbine will work, especially at the high level of execution. Hastelloy as turbine wheel material should work proprely. Pay attention to narrow gaps between the turbine wheel and the NGV as well as between the compressor and cover

kappi

 

[kappi]

a video of my Mirage III , powered by an KJ 66 -enjoy it!

 

[Blobber]

a video of my Mirage III , powered by an KJ 66 -enjoy it!

Nice video, it has a good turn of speed and looks almost like a real jet when flying!

 

[galooph]

You must need excellent spatial awareness to fly that – not easy!