Author Topic: Fuel injection systems  (Read 54923 times)

Offline Roger B

  • Global Moderator
  • Full Member
  • *****
  • Posts: 4086
  • Switzerland
Re: Fuel injection systems
« Reply #240 on: October 24, 2019, 07:51:15 PM »
I fitted the throttle and sorted out a control linkage.


And then as there was fuel in the tank even though it was in the cellar I connected up the battery and gave it a quick try  :whoohoo:


It's currently running on cooking alcohol so it doesn't smell too bad  ::)

Next step make some more needles for the diesel injectors.
Best regards

Roger

Offline Roger B

  • Global Moderator
  • Full Member
  • *****
  • Posts: 4086
  • Switzerland
Re: Fuel injection systems
« Reply #241 on: November 03, 2019, 07:05:42 AM »
On with the needles. I followed the procedure I used before to grind the first needle without problems but the next two snapped during grinding of the shank  :(  After some thought I realised that I had hardened too much of the needle so I was holding a hardened piece in the chuck. The increased stiffness caused more vibration (as shown by the chatter marks on the grinding wheel) and the brittleness resulted in the needle breaking.
I tried again just hardening 20mm (as far as I could judge) and this looks to be successful  :)
Best regards

Roger

Offline steamer

  • Global Moderator
  • Full Member
  • *****
  • Posts: 10414
  • Central Massachusetts, USA
Re: Fuel injection systems
« Reply #242 on: November 03, 2019, 11:25:49 AM »
I like it Roger!....Still watching!

Try drawing the temper just a touch before grinding....back to straw perhaps...you'll need a tray with some brass chips or sand to do this....finicky little part!

Dave
"Mister M'Andrew, don't you think steam spoils romance at sea?"
Damned ijjit!

Offline Florian Eberhard

  • Full Member
  • ****
  • Posts: 434
Re: Fuel injection systems
« Reply #243 on: November 03, 2019, 12:13:59 PM »
Hey Roger

Those are some very delicate needles to grind. I have been playing around with a small grinding attachment quite a few years ago.
A few things I learned was:
- You need to dress the grinding wheel in the lathe, using a fixed dressing tool. (not hand guided!) I couldn't see that step - but you may have done this? (If not, RDG tools sells very affordable diamond dressers)
- If you reduce the contact area to approx.  1/3 of the wheel width, you will have less deflection.  This can be done by bevelling the rear side of the wheel.
- If you need to remove "lots of material", you should angle the dremel so only the edge of the wheel is cutting. This also reduces the cutting forces in general but concentrates them to one point which is more effective.
- Always dress before the finishing pass
- I suspect this pink wheel is from proxxon? You should use them only for rough grinding. The green silicon carbide wheels work better for finsihing.
- For a very good surface finish, you can use wd40 on the workpiece (applied with a brush or so)

Just as a hint: The elastic modulus and therefore rigidity of steel doesn't change if you harden steel. All that happens is that the "resistance against plastic deformation" is increased.
Practically it means, that hardened steel will not  have any (or barely any) plastic deformation, all there is left is the elastic deformation which is reversible.

Offline steamer

  • Global Moderator
  • Full Member
  • *****
  • Posts: 10414
  • Central Massachusetts, USA
Re: Fuel injection systems
« Reply #244 on: November 03, 2019, 12:23:57 PM »
Florian is quite correct about the stiffness of steel and dressing the wheel.     You will be able to get better finish with a hardened part vs a soft part.    That said....if it's full hard it will be very brittle, and if your wheel isn't dressed, it will have 1 section of the wheel hitting it every revolution.....and that impact could break a full hard needle......definitely dress the wheel every time it's mounted

Dave
"Mister M'Andrew, don't you think steam spoils romance at sea?"
Damned ijjit!

Offline Roger B

  • Global Moderator
  • Full Member
  • *****
  • Posts: 4086
  • Switzerland
Re: Fuel injection systems
« Reply #245 on: November 03, 2019, 01:05:36 PM »
Thank you both for your thoughts and interest  :)

When I start the setup I grind the point first with a short overhang which dresses the wheel quite well.

I am increasing the depth of cut by 0.025mm and running backwards and forwards until there are no more sparks then going in another 0.025mm. I increase the depth when the wheel is at the chuck end and things are more rigid and then work out towards the point which I hope reduces the cutting load.

In my initial trials I found that the pink Proxxon wheels gave a better finish than the grey (green?) ones. Maybe I need to revisit this  :thinking: As a final step I polish the ground surfaces with a Proxxon rubber polishing drum.

As ever lots to learn and think about  ::)
Best regards

Roger

Offline Roger B

  • Global Moderator
  • Full Member
  • *****
  • Posts: 4086
  • Switzerland
Re: Fuel injection systems
« Reply #246 on: November 05, 2019, 05:40:00 PM »
The other end suffered from severe vibration but didn't break as the part in the chuck was not hardened. As a trial I held a small offcut of hardwood against the point and the vibration stopped and the surface finish improved significantly. As both Dave and Florian explained the stiffness does not change with hardening, just the likelihood of it breaking  :headscratch:

I was able to finish grinding both ends and moved onto lapping the shanks.

During the rather repetitive lapping process I realised that I had a traveling steady that Jo made to which I could fit some form of spring loaded support for the point of the needle and then use the top slide for moving the grinder. The reduced shank does not have to be that precise, It's the point that has to be concentric with the lapped shank.
Best regards

Roger

Offline Roger B

  • Global Moderator
  • Full Member
  • *****
  • Posts: 4086
  • Switzerland
Re: Fuel injection systems
« Reply #247 on: November 09, 2019, 03:10:25 PM »
I finally ended up with three nozzles with matched needles, 0.2mm, 0.35mm and 0.5mm so I can do some bench testing.

As the weather wasn't too bad I opened up the R&D department to try the slide throttle properly.


I ended up with a similar result to that from the large slide carburettor of around 120W~(~11A at 11V) however as before the cooling system is struggling. Once again I didn't seem to need to adjust the pump stroke from full load to idle  :headscratch:
Best regards

Roger

Offline Admiral_dk

  • Full Member
  • ****
  • Posts: 1498
  • Søften - Denmark
Re: Fuel injection systems
« Reply #248 on: November 09, 2019, 08:53:11 PM »
Sounds like it is running very evenly and well Roger  :ThumbsUp:
What kind of fuel was this test run on ?

Offline Roger B

  • Global Moderator
  • Full Member
  • *****
  • Posts: 4086
  • Switzerland
Re: Fuel injection systems
« Reply #249 on: November 11, 2019, 01:54:07 PM »
Thank you Per.
This trial and the two cylinder engine trial both used Alkylate fuel. I have no idea if it produces more or less power than cooking alcohol  :headscratch: I tend to use alcohol for testing the injection systems in the cellar as a hangover from when I ran the engines on normal pump petrol and wanted to minimise the smell.
Best regards

Roger

Offline Ye-Ole Steam Dude

  • Full Member
  • ****
  • Posts: 1535
  • Deep East Texas on Sam Rayburn Lake
Re: Fuel injection systems
« Reply #250 on: November 11, 2019, 03:19:35 PM »
Hello Roger,

That is just too cool....electric start and all the electronics  :cartwheel: and to top it all off, a really smooth running engine. Great job on a very nice looking engine.  :ThumbsUp:

Have a great day,
Thomas

Offline Roger B

  • Global Moderator
  • Full Member
  • *****
  • Posts: 4086
  • Switzerland
Re: Fuel injection systems
« Reply #251 on: November 16, 2019, 07:31:47 AM »
Thank you Thomas  :)

I started to re assemble the injector test rig but had problems getting a seal to the injector. On close inspection the thread on the union screw seemed worn/tapered. The union cone itself seemed ok as did the thread in the body. I'm not sure if this was a bad thread from the start or if it had worn with frequent dismantling and re assembly  :headscratch: The guide bush I use for cutting the cone in the nozzles was also a loose fit in the body so I decided to remake both parts and case harden them.

When I removed the pipe from the pump I saw that the thread on the delivery valve was also damaged  ::) This is a problem I have found with some of the Holzapfel unions. The top of the thread is not relieved and when tightened the thread spreads damaging the thread in the nut which then damages the union. The Regner unions are relieved. I made a replacement from a Regner union which meant I had to lengthen the thread a couple of mm.
Best regards

Roger

Online Vixen

  • Full Member
  • ****
  • Posts: 1383
  • Hampshire UK
Re: Fuel injection systems
« Reply #252 on: November 16, 2019, 12:37:47 PM »
Hi Roger,

I've been quietly following your fuel injection journey through all it's twists and turns. I and am sure everyone else were pleased when you got the engine to run so well. I must congratulate you on your persistence and dedication to the task. So many remakes.

I must admit that along the way I have lost track of the overall system design. Is it possible to post a diagram of your fuel injection system.

Cheers    :wine1: :wine1:

Mike
It is the journey that matters, not the destination

Offline Roger B

  • Global Moderator
  • Full Member
  • *****
  • Posts: 4086
  • Switzerland
Re: Fuel injection systems
« Reply #253 on: November 17, 2019, 11:04:20 AM »
I was unable to remove the old nipple from the pipe to replace the screw and had to make a replacement for that as well. The new nipple needs to be silver soldered on and then back to testing  ::)
Best regards

Roger

Offline Roger B

  • Global Moderator
  • Full Member
  • *****
  • Posts: 4086
  • Switzerland
Re: Fuel injection systems
« Reply #254 on: November 17, 2019, 11:10:16 AM »
Thank you Mike  :)

I will try and summarise my progress to date (this is also useful for me as I don’t have to keep digging back through what is becoming rather a long thread).
When I looked at the list of engines I would like to make most of them had some form of fuel injection so I decided I needed to do some experimentation. At the start I rather naively thought that a piece of ground silver steel in a reamed hole would be sufficient. This is obviously not the case and I moved on with learning how to lap internally and externally and then how to ream properly which required much higher speeds and lower feeds than I imagined along with the benefits of a rather expensive small floating reamer holder. After some problems with a piece of trilobed silver steel I moved onto using commercial pin gauges for the pump plunger.

First system

This was always intended to be for petrol and is fitted to my 25cc horizontal engine. The pump is 2mm bore and variable stroke, 0-2mm, driven by an excentric running at camshaft speed. The stroke is adjusted by a wedge and injection is timed to take place when the inlet valve is open. This initially ran with a needle valve injector but there was too much leakage along the needle so I tried a mushroom valve design which seems to work well.

At the start the pump was gravity fed but operation was rather inconsistent and bubbles appeared in the inlet pipe. This was due to the petrol vaporising as the plunged moved back and was solved, as in full sized mechanical petrol injection systems, by pressurising the fuel feed. I tried an automotive fuel pump but this wasn’t happy with the low flow rates so I am now using a model fuel transfer pump running at 5V instead of 12V and a relief valve that opens at about 0.8 Bar.

The first trials used a barrel type carb as the throttle and then I made a slide throttle which is now in use.

I had always planned to link the pump stroke adjustment to the throttle but that has so far not been necessary.

Second system

This is intended to be a full diesel so a different type of control is required to ensure a constant timing point for the start of injection. I decided to follow full size practice and ground a 4mm pitch helix in the plunger. I tried this successfully on the horizontal engine driven by the excentric. The helix is rotated by a rack and a wide pinion which drives a rectangular foot on the plunger.

For the diesel the pump is driven by a cam giving a short injection period, around 10°, which results in some high loads on the cam follower.

I have tried needle valve type injectors with 0.2mm, 0.35mm and 0.5mm nozzles as well as two sizes of mushroom valve injector with limited success. For the first trials the compression ratio was too low as I had not made enough allowance for the clearance between the piston and the head. With a compression of around 30 Bar I was getting combustion but I think the fuel was being ‘carburated’ as it passed by the piston.

Test system

To allow me to better understand the injectors I made a hand test pump to the same design as the first system. This immediately showed that my injectors were not performing as they should and will require various reworkings.

I have attached some pictures as well as my current drawings (makes a good backup) and will be happy to try and clarify any points.
Best regards

Roger