Austin 850 performance conversion

[DomMINIque]

3) Cylinder volume from the crown to the top of the block = area x measured height (in other words 1/2 bore squared x PI x measured height). For this measurement the head obviously has to be off the block so one can take measurement. I haven't done this on my engine because I use my car as a daily driver. I've taken the measurement on another standard 848cc engine. I'm hoping this will be close enough. In fact if anyone thinks this is a mistake and I should rather take the measurement on the actual motor, I'll have to do that. For the time being, here are some measurement pics

This measurement was 0.017" which is 0.4318 mm. Using the formula above, the volume above the crown up to the top of the block is 1.354 cc

I don't have any experience in engine building so this comes with warning.

But I don't think it would be necessary to actually change to using the actual head as I don't believe the differences between parts would make such a significant difference even if you have to consider manufacturing tolerances into account.

Based on your calculations you've got 1.354cc.

If you do this these are the results:

Lets assume that your motor currently being used as a daily has a slightly greater crown to deck height - lets use 0.5mm for arguments sake. This is 13.6% difference in measurement. This is a rather large difference in terms of engine tolerances from my limited knowledge.

(0.5 x 62.94mm)^2 x PI x 0.5 = 1.555cc

In this case you end up with a 12.92% increase in volume of the area.
Already here we can see that this is a exponential function based on the increase in measurement results in a non equal increase in volume.

Nevertheless we'll test it in the other direction as well.

In a hypothetical situation your daily uses pistons that were made on a Monday/Friday that are 0.1 mm longer than usual resulting in a 0.3318mm crown to deck height. This is a 23.15% decrease in actual measurement.

(0.5 x 62.94mm)^2 x PI x 0.3318 = 1.032cc
This results in 23.78% decrease in volume. Again illustrating the exponential effect but this time due to the fact that it is in the opposite direction the opposite result occurring. Some maths here that everyone always ask when they will actually use this in real life bla bla. (For anyone that is interested go and have a look at the properties of an exponential function)

In summary (and some plain English) I realistically, even with the mini's strange build history, I don't think your measurement between your actual motor and the other you've used to take measurements from is going to be 20% different.
And even if it is so significantly different the result would only really be 0.3cc in error (at a 25% error margin).

This works out to a total of (0.3/848) 0.035% of the total volume. And I don't think this would even affect your compression ratio calculation by even a value of 0.1.

A whole long story to basically say that I personally think the measurement you've taken for this is sufficient and it wouldn't be worth the effort stripping the other motor just for this right now.

[abri]

So, it's been a long time since I've done an update on my 850 performance conversion build.

Last time I left it at some measurements that I took. These went into a compression ratio calculator and the aim was to up the compression ratio slightly, from 8.3 to 9:1. These calculators are very handy but unfortunately I can't seem to paste a screenshot of mine in here.

Anyway, to take a step back, I started with a 12G202 cylinder head which would have come standard on the SA 998 cc engines as well as on the 997 Cooper. This is not the best head to use for this sort of project because there is a much better flowing one with less shrouded valves called the 12G295, which came a bit later and was fitted to the 998 Cooper. The 202 heads are obviously much easier to find and command less financial layout. I happened to have one that I bought a few years ago for a very good price, so that's what I used. When I bought the head it looked recently redone, so I was under the impression it would just need to be skimmed. Incidentally the 295 head has large chambers so that would also need significant skimming to up the CR for use on an 850.

So, how much can you skim off a cylinder head before ending up hitting a water jacket and having an expensive paperweight? I asked some advice. A good place to start is to measure the thickness of the head. Obviously this will give you an idea of how much (if any) has been skimmed off already. Secondly, measure the depth down the oil holes from the top of the head. Take the biggest of the oil hole depth readings and subtract that off the head thickness. What you are left with is an estimate/guide of how much "meat" is on the head that can hypothetically be removed (this assumes the spot that you measured is the lowest hollow point in the casting). Vizard recommends adding a certain "reserve" to this, to be safe from over heating (I think).

In my case the numbers stacked up like this:
Head thickness - 2.746" (this means the head had been given a 4thou lick. Standard would have been 2.750)
Depth down oil hole - 2.585"
Meat below galleries = 2.746 - 2.585 = 0.161

Subtract from this Vizard's reserve of 0.080 and the max left to skim is 0.081"

The CR calculator told me that in order to get a CR of 9:1, I had to have chamber volume of 20.8cc. Seeing as you're starting with 26.1cc on a standard 12G202 head, that is quite a large reduction in the chamber volume, especially since the chambers are quite narrow (ie. you have to skim more to reduce the volume).

Here is a picture of the combustion chamber on the 202 head, quite narrow and deep.


So I clamped the head in a vice, making sure that is was perfectly level, turned a spark plug into one of the combustion chambers, poured 20.8cc of engine cleaner into a burette and slowly drained it out into the chamber. 

It ended up looking like this


I then attempted to measure the depth from the face of the head to the fluid level and found it was pretty much spot on 0.080"



I repeated this on another chamber and got a very similar result, so I was happy that I could get the 0.080" skimmed off and end up very close to the 9:1 aim.

I took the head in for machining but was soon told that the valve seats had been cut badly and would need to be redone and matched to the valves. This unfortunately ended up recessing the valves somewhat and throwing off my meticulous calculations and measurements. Long story short, when I got the head back, 80 thou had been machined off as requested but the chamber volumes were 22.5 cc instead of 20.8, so this meant I had a CR of just above 8.5. This is not ideal but seeing as the car is just a daily driver I guess it will be ok.

More in a bit....

[abri]

I'll let some pictures do more of the talking from here...

Head was received back and chamber volumes found to be 22.5 cc which is unfortunate, but not the end of the world.

As I said before, I went for the Minispares "Performance" Small Bore head gasket. I removed the head and was happy to find the bores looking excellent with no sign of wear, no ridges, all good.

Here is the head gasket



And here is it on the block after the block had been thoroughly cleaned and all the studs turned in all the way nice and tight.



I was contemplating leaving the head in its original maroon paint as would have been standard for the 998 engine that it came from, but eventually decided it would look nicer if it was green (surprize!   ).


With the manifold off and plenty of space behind the head, I checked whether the twin H1 carbs on the Sprite manifold that I had would fit. Sadly there wasn't enough clearance, so a search for a better manifold was started. 

I asked on the Mk1 Forum whether anyone had something lying around that I could buy off them. The helpful folks on there quickly responded and pretty soon I had some options to choose from. The first option was an Aquaplane manifold like the one used here...



I was on the verge of buying that when Mr Forster of Mk1 Forum responded and said he thinks he has a tubular Downton manifold to spare. A deal was done at a very reasonable price without me seeing what the manifold looked like. I had only seen this Downton manifold  setup before (pic below), but wasn't sure what was on its way to me. Luckily the manifold I bought came with free delivery to SA as the Forster family were on their way to visit friends in SA. Small mini world 



The manifold that arrived looked similar to the design Maniflow uses these days, so its almost as if one can see the continuity (pics later).
I had in the mean time ordered a rebuild kit for the H1 carbs from Burlen Fuel Systems in the UK and set about cleaning and rebuilding the set of carbs . 

The carbs started out like this (on the Sprite manifold here):


More in a bit...

[Vinay-RS]

This really has the feel of a story. With proper cliffhangers

Seriously though, lovely work on this build. I think you and Scott need to have a race once he has built his 850 racer.

[Jared Mk3]

Looking good Abri! You're lucky to have gotten that head gasket - been out of stock for quite a while now at MS . . .

[abri]

Thanks, I've had the gasket for a few years. I've had the plan of doing this conversion for a long time but other priorities have always got in the way.

Anyway, back to the build...

So to illustrate the problem I had with the carb choke levers clearing the bulkhead cross member, below are two photos taken from more or less the same angle. The top one shows the carbs on their original Sprite manifold test fitted to a head on a bench.


The second one shows the same carbs and the same head but with the Downton manifold fitted. Notice the angle of the measuring tape and the distance between it and the bottom of the float bowl.


I test fitted this in the engine bay and the choke lever arms cleared the bulkhead cross member - only just!! But I was very happy with this because it meant I could actually use the manifold. And I was chuffed at having a 60s tuning company part on my car

So with the carbs rebuilt and fitted to the manifold, everything was test fitted to the loose head to make sure it goes together well


This exhaust manifold is from an MG1100 I believe. As far as I know they were the same as the Cooper 3-1 manifolds in terms of the diameter of the pipes. The Cooper S manifold that looks very similar has thicker diameter pipes (correct me if I have that wrong). The downpipe is certainly thicker on the S manifold. On this one you can see the weld where I had a downpipe welded on in the same shape as the Cooper manifold. This pipe is 38mm diameter on mine. 

The next challenge was what to do for the air filter system. I know most people fit the wire mesh "pancake" filters to these carbs. That's what came on the Sprite that was fitted with these carbs as far as I know. However, Vizard suggests (and other experienced people I spoke to agree) that the pancake filters are not the greatest when it comes to flow. Vizard found that K&N filters flow better than just about anything and continue to perform even when covered in dirt. My aim was performance but also proper filtering since my car is used as a daily and gets its fair share of dust and sand blown around the engine bay. So I had my mind set on a K&N filter as the ideal. I could fit two round filters or one oval one to cover both carbs. The issue was SPACE!

This setup from Vizard's book stuck in my mind and gave me the idea for mine



So I went to Masterparts and after looking at quite a few options I ended up with this



The next task was to make up a filter box to fit this to the carbs. Luckily a friend with access to CAD tools, a supply of aluminium and people who can cut and bend it came to the rescue. Before that though, he assisted me with ideas, drawings and measurements and was kind enough to have a template drawn up and cut out which I could test fit and measure for any adjustments before the final version was made (thanks Shawn!!)
 
Here is the template being test fitted: 


That was very close but with the filter in place I could see that it was not going to work without some afjustments



With the filter in place, the back plate would have to move a bit upward hold the filter in a position where it would fit. The measurements were taken, the adjustments made and the final version looked like this:


The two plates are bent to keep the filter in place at the top and bottom. Four stainless steel bolts hold the two plates together tightly on the filter. The bolts are positioned to keep the filter from moving out of position. 

The filter setup as it is, is not ideal for flow because the angle into the carb throats is basically 90 degrees. SU carbs like a ram tube that creates a venturi/funneling effect into the throat - that's why BMC/Leyland put elbows on them between the air filter box and the carb throttle body. Vizard also conclusively shows in his book flow bench results of different ram tube shapes and how much better the carbs perform with them. 


On my to-do list for this project then is having some stub stacks/short ram tubes made up that can fit inside my filter box without being too close to the cover plate. 

More updates to follow when I've uploaded more recent pics to Flickr

[Vinay-RS]

How would you be able to make the ram tubes? Just curious about the forming process. Would they be made by welding and bending aluminium tubes?

[abri]

How would you be able to make the ram tubes? Just curious about the forming process. Would they be made by welding and bending aluminium tubes?

I've chatted to Harley about it. I'm going to go and see him soon to confirm the final design, give him a carb to measure, etc. I think he's going to form them from blocks of whatever alloy. I think I will go for the flat ones with as big a radius as I can fit in there (similar to the bottom right pic in Vizard's book above). Either that or some elbows to fit the H1 carbs and lift the filter box to a more standard position similar to the BMC twin carb air filter box. I'm going to chat to Dennis about which is best.

[abri]

Time for a small update, it's been a while. 

The ram tubes have been completed and fitted to the car. I can't really say they've had a major effect on performance apart from moving the rev range where I experience the strongest pull slightly lower down. That's a good thing I guess. 

So, to take a step back, with my last update I mentioned that I'd be going with the short tubes because they'd need to fit in the air filter box. It just so happened that Vizard's tests showed that the short rounded tube is also one of the better ones for flow. 

I contacted a local engineer, took the air filter box as well as a spare throttle body to him for some measurements. We came up with the following drawings.



The the plan was to fit the ram tubes to the throttle bodies with bolts that don't come all the way through them. This was done in order to keep the surface of the mouth as smooth as possible. As you can also see, because of the shape of the filter box, the ram tubes could unfortunately not be made perfectly round. We needed to leave some space for the filter between the ram tube and the edge of filter box.

The end product looked like this



Here they are fitted to the carbs with the back plate of the filter box in place.




And finally, a shot down the throat of one of the carbs:
As you can see there is a bit of gasket sticking out and a small mismatch in the diameter. The final step before putting it on a dyno will be to clean this up with a dremel so that there is a seamless flow.


After this I have been trying different options to get the fueling and mixture correct. I've been struggling to be honest. I started out with some YY needles, recommended by someone who runs a similar setup in the UK. The car ran ok with these and accelerated fairly well at speed but was very sluggish off the line. I then tried some EB needles (a bit richer) which improved it a bit. My main issue is that if I adjust the mixture so that the sluggishness on pull away disappears, the mixture is way too rich at idle. Since I spend quite a bit of time in traffic, a mixture that's too rich at idle will quickly foul the plugs with soot. This has happened on two or three occasions, causing a misfire (not just with the EB needles). Lately I have settled on M needles (richer than EB). These needles are the exact same diameter at the section that's working at idle, so the idea was to adjust them so the mix is ok at idle but then to give it more fuel quicker to get off the line. With this I obviously assumed that it was it was a lean mix that was causing the sluggishness off the line. The sluggishness was even. It was not spluttering, so I think I was right with this assumption (and adjusting the mix richer reduced/removed the sluggishness).

Anyway, I also played around with different types of oil in the dashpots and found that heavier oil gave better results. So I'm currently running 20W50 engine oil. I did try the approved SU damper oil which appeared a bit thinner as well as an oil supplied by a local SU expert which was very thin. 

With the M needles the car runs very nicely at speed. The sluggishness off the line is not completely sorted out however, because I've had to lean the mixture out in order to be able to idle without fouling the plugs. At the moment it's still a bit rich at idle - so much so that I haven't been able to get a perfectly smooth idle. The tinkering continues...That's what you get for not keeping original parts on the car

The final setup looks like this with a choke cable recently hooked up as well.



And on the inside it looks virtually identical to the original



I have to say that it runs very well and would easily run to max on the speedometer assuming the little 850 won't go bang. I need to fit a rev counter as well

[minitim]

Lovely write-up Abri!. Ready for the track soon