Category Archives: New posts

A carburation diversion – an atomisation exploration

It’s too long since the last post but there’s been a lot going on. Racing, crashing and children with chicken pox.

So, I had the Super Squirrel on the dyno, which showed that I was down on power by almost 2 bhp from the best of my methanol runs. I had some evidence that the reduced compression ratio, combined with an ignition retarded slightly beyond optimum was capable of giving me more revs to play with at the top end but with an ever decreasing power output. The beginning of the curve was not as strong and in fact seemed to not show useful power until a few hundred rpm higher than the dope set-up. I had taken a guess at the needle jet size as 108 which seemed to be ok as we settled on the middle needle position as the optimum position, and the final main jet was 280. I’ll come back to all this later.

So I wasn’t whooping with joy, but neither was I crying in my tea. This is what development is.

After another trip to the airfield for yet more AVGAS, I thought I’d attempt to mask the early season financial hemorrhage by seeing if there were any unwanted bits and pieces I could sell as I’m all out of Kidneys. I put a couple of items on ebay and a guy in a nearby town contacted me to ask about one of them and to see whether I’d got anything else. I called him and it turned out that not only did he used to race (he won a Manx Grand Prix), that also he was a two stroke fanatic and a very recent Scott owner. He also used to run a very well regarded bike dyno cell in the south east and was interested in my dyno work. He asked me a lot of questions about the dyno which I couldn’t answer and gave me cause to doubt my assumptions about the graphs. My basic understanding is that here are different types of rolling road dynos and they can be operated in different ways. The difference between those operating methods should guide the way you interpret the results… all of course mixed in with the operators skill and understanding. So, I felt like I had more questions than answers… and still do. That’s good though, It’s when things aren’t working as well as you want them and you’ve run out of questions that you’ve got problems.

With this in mind, I went back to the dyno charts. I realise that I have taken a very literal view of the charts and assumed that the x axis (revs/time etc) was representative of the position of the throttle, and therefore the position of the needle within the jet. It’s not necessarily a completely wrong assumption, and Steve who runs the dyno was very measured in the way he opened the throttle. However, I don’t absolutely know the extent to which the graphs can be interpreted as a clear representation of engine response to throttle opening and I need to gain a better idea of what I’m looking at. As part of this I also need to understand more about how this dyno works and whether I can gather any other information that will make the graphs more useful.

In the meantime though I had my first race meeting coming up imminently and I had to work out what I could do to try and get back the missing chunk of torque at low revs. I decided to map the carbs and the needles to work out the relationship between the intake aperture and the needle aperture, with the eventual switch to the main jet. I can say 1/8 pilot, 1/4 slide cutaway then up to 3/4 needle and then main jet, but I have no idea what is actually the case in any given machine. How does that relationship actually (and measurably) manifest itself? For a start, I’ve never understood why the needle is a constant taper when the rate of increase of aperture area is not constant for any incremental lift. I expect that it’s simply a compromise born a need for manufacturing simplicity but I don’t know.

I started to look into it and then got carried away…maybe a colossal waste of time but certainly interesting!

I saw pretty quickly (after having made a calculation spreadsheet) that the 1/8,1/4,3/4 guides refer to the aperture area, and not the lift. Obvious that it should be, but I can’t remember ever seeing it written down. Working out the area of a segment when you know the radius of the aperture and the vertical height enabled me to start to put a picture together.

I measured the position of the needle (at #3 position) in relation to the top of the needle jet at the throttle closed position and then measured the corresponding point on the needle at the calculated lifts corresponding to the 1/4,1/2,and 3/4 aperture area throttle positions to give results throughout the needle range and show the transition to the main jet. I accept the possibility that there will be some inaccuracies in my results as I’m only using my eye and a vernier, and using a fine-liner to mark the needle (I don’t have an inspection microscope like my dad!) but I think it’s close enough to show something useful.

The following are the results from my investigations showing the information for one of my 1″ bore type 76 Amal carbs detailing the main jet, needle jet and needle position.

Needle jet fuel info-108-RHMtwin-needlepos4
Needle jet fuel info-108-RHMtwin-needlepos3
Needle jet fuel info-106-RHMtwin-needlepos4
Needle jet fuel info-106-RHMtwin-needlepos3

I also created a set of results for the Standard Scott type 206 carburettor, but these are based on one main assumption: that the relative positions of the needle jet and needle are the same as a type 76. They use the same needle, I think, so I can’t see how they wouldn’t be.
The Scott type 206 in standard setting has a 1 1/16″ aperture, runs a 106 needle jet and a 170 main jet. Some people seem to be increasing the main jet size up to 200 nowadays but I have not any information about the individual experiences that have led to this. I did the calcs for the standard set-up.
So, with my standard disclaimers in place!
Needle jet fuel info-Standard_Scott-106-pos3-170main

Assuming (dangerous I know) that this is representative of a correct(?) carburation relationship, what’s interesting to me is the comparison between the annular (or probably crescent shaped if the needle is against the side of the jet) aperture of the needle jet with the needle in it and the single round aperture of the main jet. Looking at the figures you can see that on my setup, the ‘dyno assessed to be appropriate’-main jet aperture is far smaller than the needle jet aperture at 3/4 opening. It was my desire to have a control result that led me to profile the standard Scott setup. On this (needle at#3), with a 170 main jet, the needle jet aperture is larger than the main jet even at 1/2 throttle.

One possibility from all this that the renolds number involved with the greater wall surface area of an annular aperture means that it has to be of a greater surface area than a single aperture to flow the same size. Another is that the needle aperture’s job is different to that of the main jet and that the needle jet/needles job of metering to the airflow at(or near) the top of the emulsion tube, requires a range of surface areas over the throttle aperture changes that are linked to the response of that fuel to a given low pressure area over the emulsion tube. This may mean that the needle/needle jet surface area simply has to be larger than the main jet at throttle apertures where you would expect the main jet to exceed the needle/needle jet aperture simply from looking at its cross sectional area.

It’s probably of limited value to compare the carburettor settings of the Standard Scott to my own (other than for interests sake) since most of the conditions are different. A Standard Scott will would have a different gas speed profile over the rev range through its single 1 1/16″ carb than mine through my twin 1″ carbs. As standard it would probably also change far less throughout the rev range being high already at low revs as the carburettor is far smaller than the inlet port area. A standard iron block I have here shows 6 x inlet ports at 19mm x 14.5mm (sorry about metric but I find it easier for sectional area stuff) which gives ~16.5cm². The throat area of a standard 1 1/16″ 206 is 5.72cm². That makes the carburettor aperture just around 35% of the inlet port area. Even taking into account the flow disturbances of the bridges, that’s a big difference. It’s probably good for low engine speed pick up, though not great for breathing at higher revs. That’s why Scott’s respond so well to inlet work to the carb and inlet tract.

My carb area is actually about 95% of my cylinder inlet port. I’ve a smaller but better flowing inlet port(s) with more carburettor aperture. I would therefore expect my gas speed over the emulsion tube to be lower, at lower revs. I may therefore require a bigger needle jet simply to give more fuel surface area to lift at these revs. However, with a straight taper, would that make me rich further up the range when the gas speed is higher?

I feel glad that I’ve gone into this, even though I know that you can tune effectively simply by changing bits until it’s right and not trying to analyse the workings of a carburettor. Certainly I’ve no solid conclusions to draw and I can’t be certain of my absolute accuracy but lots of interesting relationships and patterns have emerged that I think are valid and I hope will enable me to better develop my Scott.

I’m always open to constructive comment!

Dyno chart analysis

Andy from Performance Engineering has sent me the files from the dyno runs I’ve done from the last couple of years and with software downloaded from Winpep I am able to view all of them to compare.
I’ve been looking forward to this because I wanted to see where I could maybe make gains by modifying the needle. It’s very interesting to see the comparisons and to consider the next steps.

These are the most interesting charts, torque and power:
To be clear, the settings are as follows (a bit mixed up on the graph notes)

Run #17: Twin carb set up (avgas), High comp head (32cc)
Ign set at 22degs, main jet 280, #4 needle jet position (5bottom), 1/4t out on pilot air screw
Run #18: Same as 17 but the second run.
Run# 19: Same as above except needle now raised to #3 position (middle)
Run #28: Same as above except ign set 23.5, needle on middle position #3.

dyno graph comparison chart torque
dyno graph comparison chart power

You can see how 17 and 18 really want to rev further, and that’s with the needle on position #4. You can also see that they must be too rich at the beginning, as there’s no curve taken. I’m guessing Steve must have felt it flat and simply not sampled it at the beginning.
So it looks to me that the needle could be reduced in diameter in the final section to enable a bit more richness at the top end without compromising the bottom. That should give me an extra five hundred revs according to the chart. That should give me the best of the start of #19 and the end of #17/#18. I can work the amount out simply by checking the difference in the needle within the pitch of a needle adjustment notch.
The timing change seems to have a decent effect in increasing peak torque across the range, so I’ll leave that where it is.
It’s interesting that the needle position change seems to have had an effect within an area that you’d expect to be almost beyond its influence, as I understand it. I need to check the needle jet and needle at that position to see whether it’s restricting the main jet. You’d think that it could be, looking at the graph.
As for the beginning and my lost 400revs, I’m not exactly sure. Part of me thinks that it might be weak because the fuel is too low in the needle jet as I’ve set it to not drip out the pilot air screw assembly. As the carb is at an angle, this obviously reduces the level in the needle jet. This theory is slightly supported by the fact that the very first section (around 2750 revs) was a little stronger with the 300 main jet and needle at the highest position.
However, the ‘technical guy’ from the Amal stand at Stafford said that the float level was set at 0.9″ from the top face of the float banjo. Extrapolating that into the carb means that this works out to be within the thread for the needle jet in the jet block. That’s not very far up the needle jet anyway.
Later carbs made for two strokes had the rear part of the emulsion tube cut away as this created a situation where more of a low pressure area was created to lift the fuel from the jet. Apparently (Amal guy) this was done when the higher performance two strokes came along. It may be that the response time was shortened and at higher revs this became important… I just don’t know. The other ‘known unknown’ is that I think the bellmouths aren’t made with the requirements of gas flow as the primary design factor. I suspected that they wouldn’t be perfect but the way they are made leaves quite a significant ridge around the inside against the edge of the carb inlet. Not ideal for laminar flow as I understand. It may be that this, along with the less than ideal brief radius at the end of the bellmouth is responsible for a meaningful restiction to flow. I didn’t take into account in any mathematical way the extra venturi wall area when I originally planned my conversion to the twin 1″ carbs, but the overall increase in aperture is around 10 to 15%. If I’ve got bad turbulence into the carb then that might cause problems..also since the emulsion tube is right at the bottom of the venturi, it could be that disruption to the flow entering the carburettor is effecting the flow over the emulsion tube.
I think I’m going to try to temporarily effect a radiused entry using silicone sealant and see how it works. The radiused bellmouth I’m going to try and do using epoxy resin, just to test. I’m probably not going to know, but I will have the bike back on the dyno at some point.
Longer term, I’m probably going to file both carbs out to 1 1/16″ (around the top and sides) to increase flow, but that’s not for now… oh and make some decent bellmouths.

Stafford Show visit – April 2015

I hadn’t resolved to go to Stafford until the day before, but I knew that Roger was going and that it was also a good chance to catch up with people who I’ve known through my life around Scotts, but who I have seen rarely in the last few years, since I stopped building engines for a living with Roger. I also had a chance to talk to the technical guy on the Amal carburettor stand about the level the fuel should be within the emulsion tube of a type 76.
I’m constantly thinking of how to improve the power delivery right at the bottom end.
The Scott stand looked very good and had a back drop of screens with interesting photographs, including several of Scott’s in competition over the years. Sheelagh Neal had her father Ossie’s famous Scott racing outfit on the stand and there was a Reynolds special, Harry Langman’s TT racing outfit and a Sand racing Scott.

I also had the opportunity to have a look around the Bonham auction. They had a couple of quite iconic Scotts as well as a Silk. One of the Scotts was apparently a 1926 TT entry and was the first known use of the duplex frame. The other Scott is well known to me as it was the actual machine that I always thought to be the most beautiful Scott when I was very young. It is a Sprint Special and belonged to Dennis Howard, and then Glyn Chambers. It is pictured in the Jeff Clew book, ‘The Yowling Two Stroke’. A special bike I think.

My dad also had a small display of engine parts to enable people to see that there was new spares support for the marque. I think it also shows that the bikes are living and that people don’t have to fear using them.

It was a good show. A few pictures below:

A morning on the dyno with the Scott

I really enjoyed taking the Scott to the dyno yesterday. It was the first time I’d ever actually booked a proper set-up rather than a single run. I suppose cost was always an issue but I have realised how useful it is to be able to see the results of the changes you make and a proper session allows you to experiment with many different settings and see how each affects the result. It’s a bloody good way to see what affects your power output, and how. The cost is similar to me going to a trackday, and there’s no way I’d have gained the same information from that.
Of course, you need to have someone who can run the dyno and can interpret results. GT motorcycles of Plymouth have a very interesting division called ‘Performance Engineering‘ that specialise in rebuilding, restoring and tuning Honda racing bikes, and it’s these guys that run the dyno. Steve is highly knowledgeable and experienced in getting the most from an engine on the dyno. I’ve done my previous runs with him, but yesterday he was able to really show me the benefits of a dyno setup. Steves colleague Andy, another two stroke fanatic, was assisting and as the Scott’s engine note became crisper and cleaner they were both obviously enjoying the experience. Ok, so the Scott doesn’t put out the sort of horsepower as an NSR500, but it did sound so very good .. and maybe the 4.5% Castor mixture helps too!

Anyway, I realise that for many people the idea of going on a dyno seems a bit extreme but carburation and ignition set-up problems can persist for years if you don’t seize the opportunity to really get to the bottom of them. I don’t want to spend my time at a race meeting or a track day like the Beezumph constantly messing around with jets , needles and timing just to try and get somewhere close. You’ll always have a certain amount of tweaking.. but at least I feel like I am quite close with what was a completely unknown arrangement.

So, What happened?

I’d fitted the 1″ carbs with 320 main jets for a start. The needle was on the bottom groove, so all the way up. I’d figured that it might be rich but I had no desire to start from the other direction! The timing started at 22BTDC.

It was rich. Our runs followed the following order:

1: down two mainjet sizes to 300
2: down two mainjet sizes to 280
3: drop needle one notch
4: drop needle another notch
5: drop mainjet one size
6: raise mainjet one size
7: change timing to 23.5

You can see the graphs here for the final results compared to the very good curve I produced when I first set up the bike up on methanol.

Torque curve for the Super Squirrel racer.
Torque curve for the Super Squirrel racer.
horsepower curve for Super Squirrel racer
horsepower curve for Super Squirrel racer

The exhaust temperature sensors were really useful and we were able to monitor for significant changes with any modifications to setup that we made. The final high was around 530°C. Always increasing by about 30°C to 40°C the moment you shut the throttle off… interesting stuff.

Type K thermocouples read by simple meter
Type K thermocouples read by simple meter

You can see that the Torque curve starts later and then increases at a very high rate. It may be that this lost section (around 400 revs) at the beginning is retrievable by playing with cutaways, but the gain in midrange torque is very good indeed. That may be down to the improved gas flow in the head. It loses to the methanol later in the rev range, about 4000rpm. The methanol torque stays high all the way to 5000rpm where it suddenly falls off a cliff. The new curve hits a higher torque figure but tails off more noticeably. However, the fact that suddenly I started to have more revs was a bonus (as long as I can get the 400 back at the beginning!). I think Steve’s attitude was that the extra ‘over rev’ was extremely useful as the engine was able to keep pulling longer because of it.
As far as I understand, the fact that I increased the head volume, lowering the compression ratio might have created the conditions to encourage the engine to keep revving.

As the torque was lower late in the new curve the maximum power is less than before, but I’m not too concerned about that at the moment. It might be that I could do with the shape that I have created in the combustion chambers, but with a high compression ratio. It would be worth experimenting with heads sometime.. but not now. It may even be that the carb size is a little small, and I’ve never done any further tests with the exhausts but short of changing the slides to try a higher cutaway I’m now looking at actually using it! It felt really smooth and sounded great.

So, this afternoon I posted off my entry to the British Historic racing clubs second meeting of the year at Anglesey on the weekend of the 16th and 17th May. Roger is going too. I’ve never ridden Anglesey before but I’m really looking forward to it.

Test run observations

I took the Super Squirrel to my friends farm today to run it down the track, as I didn’t really want to take it to the dyno without knowing it even works. I fitted the pair of gauze covered bellmouths as the road is quite stoney and I really didn’t want any unwanted inhalations. Radiator filled, fuel filled and away it went immediately.
It sounded clean but the first thing I noticed was that it was settling to a high tickover, rather than the stop that we need it to have. I need to check the slides and pilot settings.
It measured between 270°C and 300°C on the exhaust temperature sensors, which doesn’t mean anything much at this point, but might mean more as we move toward a better set up tomorrow.
Pulling away up the hill it was getting bogged down, so I imagine that it’s rich on the needles, which are standard 276 needles raised to the top as a starting point. I don’t think I got to the mains, but we’ll check tomorrow.
So tonight I’ll tweak the bits I can and the rest will have to be done tomorrow.
It’s really a tall order to try and get the whole set-up done in a few hours but we’ll see how far we get.

Scott racer developments

One of my main aims for this year for my ’32 Scott racer was to set up the carburation properly for petrol again, having decided to move away from Methanol. Methanol worked really well, and the power characteristics really suited the three speed box. The bike was fast and really responsive to ride with the single overbored Amal type 89, but I worked out that the venturi area was less than optimum and resolved to improve that.
With the single carb, the shape and position of the torque curve was pretty much exactly what I’d expected based on the calculations I’d made using Gordon Jennings formulas relating to port time/area relationships. Basically, the you need a certain area of port available to enable adequate gas flow at a certain revs and that area increases as the revs increase as the gas has less time to pass through it. Nothing I have done over this winter should change that torque curve shape too much, as I’ve not changed port timings. Changes in exhaust pulse characteristics relating to the increase in temperature anticipated with the change to petrol will make a difference, and quite possibly not a beneficial one but we’ll see.

Vee twin manifold
Vee twin manifold
I do expect that the new fuel system, using a pair of 1″ Amal type 76 carbs on a vee type manifold, should keep that curve from tailing off quite when it does at the moment. I spent hours setting up the float heights and chamber positions so I hope it works. The extra breathing coupled with an attempt to tune the inlet lengths to work better with the slightly extended inlet timing duration I have on this engine should in theory pay off. With the single carb, you could see an inch or so ‘stand off’ of mixture blown out (and sucked back) at low revs. The longer tracts will help eliminate that. I have also spent a significant time working on the head profile to try to allow a more direct route for the flame front to move to the extremities of the combustion chamber but there’s a fair bit of finger in the air stuff… with a bit of borrowed knowledge and the rest; ideas formed through slow but incremental observation of what has already happened.

So, with a completely new fuel system, I need testing to get it right. The most straightforward way for me to do this is by getting onto the dyno that I’ve been using for my other tests so far. This time though, not to do a run but to do a full set up.

So this week, I’ve booked a morning at Alan Jeffery’s dyno in Plymouth, run by Steve from GT motorcycles who seems to spend most of his time tuning NSR 500 Honda GP bikes for people all over the place. A good man to have on-side.

I’ve tried to make sure that I’ll be ready to make the most of it. Roger sent me his block of main jets,IMG_5138 so I’ve got changes to make. I’ve fitted the exhaust temperature sensors to enable me to use that information to help establish whether we are too lean. I’ve got a couple of plugs from a different heat range and on top of that… I also picked up some AVGAS as I’ve been convinced of the advantages of this over even the higher octane rated unleaded from the forecourt.
I’ve also ordered a radiator hose connector which allows the fitting of an 1/8″ temperature sensor and at some point I’d like to fit one of these to go with my Scitsu rev counter which (although it probably needs servicing) stopped working altogether when I switched to methanol.

Hopefully, I’ll get a better result that the first dyno run I did last year. That gave a maximum torque figure of around 38ft/lbs at near to 4000 rpm and a maximum horsepower of 33 at 5000rpm.

I really want to get out to do more racing this year and I really want it to be competitive.

March 2015 update

After I decided to return to running petrol in the Super Squirrel racer, I thought that I’d re-profile the cylinder head ‘squish’ chamber, given that the compression was probably on the high side and I really wanted to concentrate on trying to create a head shape which made it easier for the flame front to spread.

The original head chamber and the beginnings of work to the other side.
The original head chamber and the beginnings of work to the other side.
The old shape worked well on methanol but I know that I am more likely to suffer from detonation with petrol and so I thought it prudent to make some changes accordingly.
Finished re-profiling and blending. I have not idea whether it'll be better but it seemed like the right thing to do.
Finished re-profiling and blending. I have not idea whether it’ll be better but it seemed like the right thing to do.
I also asked Alan if he’d braze a couple of fixing points onto my exhaust headers, to allow the installation of exhaust temperature sensors as this information may be useful in determining the correct jetting.It will be interesting to note the changes anyway.
The K type thermocouple has platinum probes which will screw into the nuts and poke through a 2.1mm hole.
The K type thermocouple has platinum probes which will screw into the nuts and poke through a 2.1mm hole.
I’m also intending to monitor the temperature underneath the spark plugs, to see what that tells me. I’d like to monitor coolant temperature also but I’ve not yet picked up a sensor for that. Whether any of this stuff gives me absolutely necessary information is questionable, but it will certainly be interesting.
I’ve also been working on the detail of the new improved twin carb set-up. I’ve not actually set them up(!) but I am almost at the point when I can put it all together.
Lots of time spent on these, scraping base flanges flat and careful flow work to match manifold.
Lots of time spent on these, scraping base flanges flat and careful flow work to match manifold.
I’ve spend quite a lot of time matching the carbs to the inlet manifold and also the flow going in. I also have to clean the jet block fuel and air apertures and work out whether I need to increase the size of the pilot jet apertures. I’m not bothered about tickover of course, but the pilot is the only source of fuel when the throttle is shut and I wonder whether, at the end of a long straight, this is necessary for cooling the piston.
A bit of cleaning and fettling for the jet blocks
A bit of cleaning and fettling for the jet blocks
Maybe it’s not, but nip ups quite often happen when you shut the throttle and it strikes me that this might be related. At any rate, the carbs were orignally destined for something like a 350 single running a one hundred and something main jet. I’m going to be running considerably more than that so I’m surmising that the pilot jet should be appropriately modified too.
I’m also planning to try different lengths of intake tract for the carbs as I extended the intake timing duration a little and it will benefit from some length and velocity on the inlet.
Otherwise, I have to do a lot of cleaning of gritty parts, and then get it back together for testing.

Dash to the North – the completion of the Silk Scott frame

I have a whiteboard in my workshop and it generally lists too many tasks that I am a reasonable amount of effort away from achieving. This last week however, I managed to get some significant ‘to do’s’ rubbed off the board.
The Silk Scott frame had been waiting at Alan Noakes’s workshop for me to come up with a plan for the final brazing solution. The lugs had been made to be brazed with a capillary fixing which really needed oxy-acetylene. I must admit that we hadn’t fixed the exact type of brazing material we were going to use and this was all part of the delay. I saw a clear weekend coming up and thought that this was my chance to push to get it finished. So the plan unfurled: first to get oxy/acetylene.
We’d had a plasma cutter at work that we’d bought from machine mart (please no comments). It was around £600 new and lasted a year and a day. The warranty department, were sympathetic but not quite to the point of being reasonable or useful, and I was told (after paying for the service) that it could be repaired for a sum of 500 and something pounds. I declined their kind offer. We’d only used it a handful of times to make register plates for woodburners and other bits and pieces. Normally around 3-4mm thick steel for a machine rated for 10mm. Anyway, since this incident I am resolved not to use them again and so was without any metal cutting equipment at work apart from grinders. So, I figured at £5 each per month for bottle hire, I could justify the oxy/acetylene. My Dad had an old portapak I could have and so I found myself at BOC sorting out an account last week.
I called Alan from the shop and enquired as to whether he’d got any prior arrangements for the weekend, and asked about rods. He didn’t and It didn’t seem that they had anything suitable, so I left with two bottles hoping that I could sort out the rest in the following few days.
I remembered a conversation with a man called Arthur Sosbe not long ago. Arthur, though now mostly retired, is a Leicester welder held in very high regard by my father. He is also a vintage motorcycle enthusiast and I believe used to race a velocette that used to lurk under a cloth in his workshop. I myself have known Arthur for many years and he has repaired many a Scott crankcase, as well as frost damaged barrels and many other fragile vintage parts.
Arthur had said that he’d naturally use silver solder, but when I quoted him a 0.010″ gap, he said that it was too wide and he suggested something else. I remembered this conversation and also that many lugged cycle frames used silver solder and I did a bit of research. A company called ‘Cupalloys’ came up as being suppliers of silver solder to model engineers, so I gave them a ring. I’d since confirmed the gap with Alan as being nearer 0.006″, which was within the capillary range for a 38% silver solder alloy which also apparently had the benefit of melting over a reasonably wide heat range, which gave scope to also create fillets.
I bought two packs of 5 rods, and at over £4 per rod, I hoped this was going to work.
So Friday came and later than I’d hoped, I packed the Moss crankcase destined for the Silk Scott into the van and I headed up country to Leicester to see Roger and to get the portapak and hoses/ regulators I needed to take up to Alan. I didn’t arrive until after 9pm but we gathered all the bits and pieces together, as well as a trophy I’d been awarded by the British Historic Racing club, the ‘Aotearoa Trophy’, gifted once to the club by a New Zealander. I believe it’s supposed to be the best performance of a 1930 and under bike. I won it last year also, and although it’s not that I actually won a race, it’s a beautiful shield with names going back many years.
We chatted, a lot, and eventually I got to bed around 1.30am.
So up and off to Alans, the offside front wheelbearing of my van starting to whine somewhat irritatingly.
I arrived at Alan’s somewhere after 10.30am and had thoughts that I might need to find a B & B for the night, as I didn’t know how fast the job would go with untested equipment and solder. I needn’t have worried because apart from Alan having to adapt some spacers to fit the crankcase, the whole process went very quickly.

Moss Scott crankcase fitted into frame - just to be sure!
Moss Scott crankcase fitted into frame – just to be sure!
we did take a few minutes to try to establish that the frame was straight prior to fixing it, and to that end I took a sight from the rear engine mounting to a straight edge laid over the bottom face of the head stock. This looked spot on, although pretty much every other tube in the middle looked like it was in a slightly different position. I realised that it was simply built this way from new and that as long as the main datum points were correct, the bits in had some ‘tolerance’. That of course is one of the joys of a low production handbuilt machine. Every one is truly different!
So onto the hot work and (making sure the detachable lugs all faced in the correct directions) using a propane torch and the oxy-acetylene, Alan soldered the new cradle assembly into the frame within a few short hours.
Alan at work 001
Alan at work 001


Nearly finished...
Nearly finished…
I was very grateful for the sandwiches made by Margaret for lunch and delighted with the way the frame came together.
with a thanks and farewell to Alan, after he’d showed me some detailed bits he’d made for Scott TT replica forks,
Scott TT replica fork crowns
Scott TT replica fork crowns
I shoveled my frame and the bits and pieces back into the van and headed back to Devon. A quick stop in at my dad’s to show him the results, but all in all a great result for the weekend.. and I was home on Saturday night.

Colin Heaths racing barrels

I emailed Colin a couple of days ago to ask him whether he knew of anyone who had cut the bridges out of the transfer successfully and he sent me this. Thanks Colin.

“Just a little more about the pictures. These are the barrels used by Martin Heath in that one glorious season when he won 12 events, including his first win. This was at Cadwell when he was left on the grid as the compression was so high the rear wheel just slid instead of turning the engine over, on the old downhill start by the timing/ commentary box. He managed to start it eventually by vaulting on from the greatest possible height and it fired up. The field were disappearing into Charlies by now, but he caught and passed them all to win. Proper Boys Own Paper stuff.

However, the point is that the barrel, as we had no other, was reclaimed by boring out the broken skirt from a damaged set and pressing in a flanged liner from the top. ( It sat in a machined recess). The machining had broken through into the jacket so good old Loctite sealed it all up and provided what I think is called a ‘wet liner’. We could not bring ourselves to put dividing bars back in the ports so made them elliptical as you see and relieved them to give the rings a softer time. You will also see that we needed to use a ‘detachable’ steel ring for the lower seal – and this located on a small step machined on the liner o/d.

Another simultaneous experiment was to make the liner full length down to the very bottom of the crankwell to see whether preventing any piston ‘rock’ would help. You can see the remains of this in the pics after it was later cut off. The rods had to be scalloped to clear the base of the liner, but we still use them to this day with no trouble.

The heads are a type you are familiar with and the drilled/tapped holes onder the dome on the exhaust side were so that we could use a head steady onto a frame cross tube. ( This was our super lightweight T45 frame, it was more of a frame steady than an engine steady. It was so light it sang like a tuning fork even after the engine was cut).”

I also sent him the pictures of my piston for his interest and to get his thoughts.

“As for the fuel pattern on the piston crown, by using a close profile high compression arrangement I expected that any ‘clean’ area on the side was trapped fuel ‘end gasses’ that got rudely pushed/ sucked out of the way before they had a chance to detonate.”

and of the photo of Martin:

I think it was taken at Mallory but expect Martin will correct me if necessary. Incidentally the silencer shown was our first effort after introduction of silencing regulations. This one is designed on the principle of ‘silencing by controlled leakage’ whereby multiple small outlets are provided under the crankcase all carefully pointed in different directions. From memory this set up has five intentional outlets including the official one. The theory is that one noise meter will have difficulty covering all directions at once. It worked well enough, for we would have been excluded on open pipes, and to my surprise the machine seemed, subjectively, to go just a little better than before.

March update: Super Squirrel carburation

The way I’d finished 2014 was with a rough set up of twin carbs on methanol. I’d modified the needles and the jets, but it was all way too rich (post meeting dyno results here) and a rush job really. The truth is, sometimes you have to push through to get something done, knowing that it might need work later. I decided that I would move away from methanol anyway after that last meeting and so it was a case of starting again.

I measured the port sizes last year and the inlet on each cylinder of the Super Squirrel is about 10.25cm². It’s actually less than standard, as we block the rear most inlet apertures up to allow us to put windows in our pistons to aid transfer flow. A Scott standard inlet gallery (cast into the skirt of the barrel) has a lot of small bridged ports which work fine in a standard road machine. In a racing bike, they are not able to flow enough gas quickly enough once the work has been done to enlarge the inlet tract in the crankcase which feeds them. This inlet tract is very restrictive in size and any serious attempts at performance usually involve some substantial work to open this up. The ‘floor’ of the tract rises up as it follows the radius of the central flywheel which is housed immediately beneath it. What we used to do was to reduce the diameter of the flywheel and weld up the bottom of the tract to enable it to be opened out. Of course such a measure means complete re-qualification of the bearing cups and barrel locations as the whole lot will be pulled in to the middle. Not a job for the faint hearted! You can of course improve performance by working on the tract without doing the floor, but it just depends how much performance you want to extract. My crankcase is one of Rogers improved castings with a better floor profile, thicker sections and far better material.

So the inlet was around 10.25cm², but I was using a single overbored Amal type 89,

Ovally bored single carb used on the Super Squirrel since 1970.
Ovally bored single carb used on the Super Squirrel since 1970.
which had been cobbled onto a Scott flange many years ago. It was about as big as you could get stuck behind the single downtube and was/is a really good carb, giving great pickup out of corners with no hesitancy, possibly down to the good gas speed due to its small size. However I think it’s undersized really as its aperture is only 8.04cm². That’s over 20% smaller than one of the cylinder aperture areas.
Since the Super Squirrel frame will not fit a larger single carb due to the frame tube being in the way, a split manifold seemed to be a good idea. Eddie Shermer made the one I have.
Twin carb manifold made by Eddie Shermer
Twin carb manifold made by Eddie Shermer

I figured that I would try and get close to the inlet gallery area with the carbs, and so settled on the idea of 2 x 1″ Amal 276 (or 76) with remote float(s) as this pairing give an area of 10.13cm². Although I don’t know what the relative flow characteristics would be between one carb or two of half the area, I guess that the flow rate might be less with two due to a greater proportion of drag from the inside of the greater surface area of carb venturi wall. However, the atomisation could be better. We’ll see. I actually had a 7/8″ Amal 276 and and 1″ 76 and was toying with the idea of getting the 276 bored out. I decide instead to keep an eye on ebay to see if something came up and thanks to someone on the Scott Owners Club forum who spotted one, I bought a singularly unhygienic 1″ Type 76 a couple of weeks ago. It had exactly the same smell as I remembered when my dad’s shed was visited by rats many years ago. I remembered they were so big that the cats used to sleep on high surfaces to avoid having to meet them. It required a lot of scrubbing to get to a point where I could inspect it…

So just the body and the jet block, well corroded in. An evening of gentle heat, wd40 and physical persuasion wrought success and now I am on the way to getting the carb setup back together.

While I was doing this, I remembered that I’d been meaning to remove the head to see if I could detect any patterns which might give some information as to how the bike was running. Initially this came out of a conversation with Ignition and bantam tuning expert, Rex Caunt at Cadwell Park last year and even though I was changing fuels I thought I’d have a look.
I was very interesting…

It could be that the clear sides of the crown indicate that I’ve achieved a functional ‘squish’ clearance preventing detonation (excellent article here). I’d really like to think that but I don’t (next day edit: I’m coming round, it might be a bit of this). It looks more like a tell tale that you wouldn’t normally see, but that my significantly over rich mixture is allowing me to see; the loss of unburnt mixture from the transfers straight owner the top corners of the piston into the exhaust. I might be wrong, but that’s my feeling.
It’s easy to see why it would happen, the transfer port is a bridged rectangular port which operates onto a deflector which is not rectangular, but is rounded at the top corners. The crankcase has stuffers cast in, which is a legacy really from the times where we weren’t running extractor exhausts. With stuffers and an extractor exhaust maybe it’s all a bit high speed into the cylinder and too much is being lost. Although I would expect to have some fresh gas returned to the cylinder prior to the closure of the exhaust port, this initial transfer gas would be short-cutting the scavenge cycle and so is not only unlikely to be returned but also not actually scavenging the cylinder of the remaining exhaust gases. What can I do if it is this? Well, Maybe some better shaping of the transfer bridge would reduce the tendency of the gas to bifurcate and disappear up and over the deflector. Maybe some mild shaping of the deflector could help. I had thought about removing the bridge but it does look like it’s too wide for the ring to cope.
I think it might be a case of planning a better cylinder and possibly looking at this for the Moss/Silk Scott racer too.

It’s all quite late now … I did spend about two hours scraping flat the exceedingly distorted flange on the new carb I bought but it’s still not done. Still, It’s all moving in the right direction.

Getting there...
Getting there…