In 1969 Gordon
Blair designed a radical new racing exhaust system for the ageing Matchless G50
engine. The design of this system was
based on data obtained from an engine simulation model, a process that
has now become the accepted norm.
Blair/Seeley G50 megaphone |
Gordon Blair
studied mechanical engineering at Queens University Belfast, completing his BSc
in 1959, a PhD in 1962 and a DSc in 1978. He returned to the Department as an academic in 1964
and eventually retired as an Emeritus Professor in 1996. Gordon always had an
interest in motorcycle racing and a great admiration for fellow Ulsterman Joe
Craig, the Chief Engineer of Norton Motorcycles. Craig designed and developed the
single-cylinder 500cc Manx Norton racing engine from 1931 to 1954, almost doubling
the power output within this period.
'The Professor' Joe Craig |
In 1965
Gordon commenced his involvement in motorcycle sport with machines and
components designed and built at Queen’s University. His work using validated
computer simulation programmes was now paying dividends in engine performance,
giving the Queen’s University team an advantage over even the factory teams.
With the outstanding Irish rider Ray McCullough working as a technician in the mechanical engineering laboratories, this was effectively the start of QUB Racing. In the late 1960s Brian Steenson joined the Department to complete a PhD under Gordon Blair. Brian had worked in the research department at BSA but was also a gifted rider in his own right.
Gordon Blair with 500cc URM |
With the outstanding Irish rider Ray McCullough working as a technician in the mechanical engineering laboratories, this was effectively the start of QUB Racing. In the late 1960s Brian Steenson joined the Department to complete a PhD under Gordon Blair. Brian had worked in the research department at BSA but was also a gifted rider in his own right.
Brian Steenson G50 Seeley |
The G50
engine was now being manufactured by Colin Seeley Racing Developments Ltd. and
was fitted to their Seeley-framed 500cc racing machines. Traditionally these
engines were fitted with a short primary-pipe and short reverse cone megaphone
exhaust system, the dimensions of which were based on the development data
recorded by the engine designer Jack Williams.
Seeley brochure showing Std G50 exhaust |
Blair’s
design was very different in that the megaphone was a very slow taper, with an
overall length more than three times that of the original Williams design.
Both designs are classed as reverse cone megaphones. It is the diffuser section that is referred to as the megaphone, whilst the short nozzle section the reverse cone.
Both designs are classed as reverse cone megaphones. It is the diffuser section that is referred to as the megaphone, whilst the short nozzle section the reverse cone.
Williams design: Primary pipe length 813mm
Megaphone
length 265mm including 25mm
reverse cone
Megaphone
outlet 98.5mm diameter
Blair design: Primary pipe length 840mm
Megaphone
length 915mm
Megaphone
outlet 86mm diameter
Blair’s
design caused some consternation when it first appeared on Brian Steenson’s
Irish Racing Motorcycles Seeley G50. Some of this was actual, some of it
perceived, but there was no doubting that the machine in his hands was very
fast indeed. Brian finished 2nd to Giacomo Agostini on the factory
MV Agusta in the 1969 Senior Ulster Grand Prix, beating all the established GP
stars convincingly in the process. It was found the engine produced more power
above 6750rpm and a considerably higher peak value than when using the standard
G50 exhaust, resulting in it being able to pull a higher final drive ratio.
This it did with an additional 200rpm giving a valuable increase in top speed. It
was also significantly more economical on fuel, resulting in a lighter initial fuel
load and less frequent fuel stops in long races. However the exhaust system
emitted a flat, almost toneless note which made the engine sound less powerful
and lower revving than the standard pipe with traditional single-cylinder rasp.
The question to be addressed is why did Blair's megaphone produce a different performance characteristic than the standard Williams G50 type system?
If we look at the diffuser section of the standard G50 megaphone, it is apparent that the diffuser angle is significant at circa 8.7 degrees. When one designs expanding ducts, there is the old advice that 'flow separation will occur if the diffuser angle exceeds 7 degrees'. Now, in practice the single 7 degree figure is not really satisfactory as the likelihood of separation of course depends on the Reynolds number of the fluid (in this case exhaust gas), however what it does illustrate is that the standard G50 megaphone will exhibit flow separation. In the case of the Blair design, the diffuser angle is a mere 1.5 degrees and as such separation is likely to be avoided.
However, flow separation is not the main factor which resulted in the change in performance at elevated engine speeds (6750rpm and above). As noted previously on the bike, the following characteristics were found with the Blair exhaust fitted:
When the pressure traces are compared, a clear impact on the in-cylinder pressure is observed; the cylinder pressure at intake valve opening with the Blair megaphone is significantly lower at 1.5bar compared to 1.9bar with the std G50 exhaust.
With the Blair megaphone the lower amplitude 'double hump' exhaust pressure trace results in a lower mean cylinder pressure during the whole of the exhaust stroke.
The old 'back of a fag packet' calculation to estimate pumping losses is that it can be approximated by the difference between the intake and exhaust pressure. In the case of the Blair exhaust system, it is the lower mean exhaust pressure which results in a drop in the pumping losses.
With a drop in pumping losses, the Blair exhaust therefore has an increase in brake torque/power and also an improvement in engine thermal efficiency (due to the reduced losses) and hence fuel economy.
The lower cylinder pressure at inlet valve opening with the Blair exhaust also has another advantage; blow back of in-cylinder residual charge into the intake runner will be reduced during the valve overlap period. In-cylinder residuals are hot and are also composed of oxygen less gas (it has been combusted). As such, the lower levels of in-cylinder residuals being pushed back into the intake port with the Blair exhaust results in a intake higher charge purity and what charge is there is also at a lower temperature. The advantage of the latter is that lower temperatures result in a higher density, which in the case of a racing engine mean a higher mass of air and fuel being induced per cycle. The result of both these factors is that the Blair exhaust will allow a greater mass flow of air and fuel per cycle into the cylinder and as such will increase the engine torque/power further.
The resultant performance of the Blair exhaust is a 3bhp increase in engine power at 7250rpm over the std G50 exhaust.
It does however need to be noted that over the lower speed operating regime (below 6500rpm) pumping losses were actually higher with Blair's megaphone compared to the std G50. This would cause a drop in torque/power at speeds below 6500rpm. A few key conclusions about what this means on a track:
.
The question to be addressed is why did Blair's megaphone produce a different performance characteristic than the standard Williams G50 type system?
If we look at the diffuser section of the standard G50 megaphone, it is apparent that the diffuser angle is significant at circa 8.7 degrees. When one designs expanding ducts, there is the old advice that 'flow separation will occur if the diffuser angle exceeds 7 degrees'. Now, in practice the single 7 degree figure is not really satisfactory as the likelihood of separation of course depends on the Reynolds number of the fluid (in this case exhaust gas), however what it does illustrate is that the standard G50 megaphone will exhibit flow separation. In the case of the Blair design, the diffuser angle is a mere 1.5 degrees and as such separation is likely to be avoided.
However, flow separation is not the main factor which resulted in the change in performance at elevated engine speeds (6750rpm and above). As noted previously on the bike, the following characteristics were found with the Blair exhaust fitted:
- Increase in power above 6750rpm
- Improved fuel economy
The characteristic that defined these changes is due to the change in exhaust pulsations and unsteady gas flow. Quite simply an improvement in both power and fuel economy is likely to come from a reduction in pumping losses. Pumping losses are the energy loss in a 4 stroke engine during the gas exchange strokes (intake and exhaust). They are a negative thing for engine performance and economy; ideally they would be zero (or even positive). To try to explain pumping losses, when you take you foot off the accelerator of you car, it is apparent that the vehicle slows down considerably; often termed 'engine braking'. It is pumping losses that have a significant contribution to this braking/negative work effect.
The exhaust pulsations are very different between the std G50 megaphone and the Blair design.
- Std G50 - A main 'single hump' pressure peak in the exhaust occurs (circa 2.3bar absolute) near bottom dead centre at the start of the exhaust stroke.
- Blair megaphone - The main exhaust pressure pulse now has a 'double hump' with significantly a lower peak pressure (circa 1.5 bar absolute)
When the pressure traces are compared, a clear impact on the in-cylinder pressure is observed; the cylinder pressure at intake valve opening with the Blair megaphone is significantly lower at 1.5bar compared to 1.9bar with the std G50 exhaust.
With the Blair megaphone the lower amplitude 'double hump' exhaust pressure trace results in a lower mean cylinder pressure during the whole of the exhaust stroke.
Log P v Log V 'Indicating' diagram showing how an increased exhaust pressure can increase pumping losses |
The old 'back of a fag packet' calculation to estimate pumping losses is that it can be approximated by the difference between the intake and exhaust pressure. In the case of the Blair exhaust system, it is the lower mean exhaust pressure which results in a drop in the pumping losses.
With a drop in pumping losses, the Blair exhaust therefore has an increase in brake torque/power and also an improvement in engine thermal efficiency (due to the reduced losses) and hence fuel economy.
The lower cylinder pressure at inlet valve opening with the Blair exhaust also has another advantage; blow back of in-cylinder residual charge into the intake runner will be reduced during the valve overlap period. In-cylinder residuals are hot and are also composed of oxygen less gas (it has been combusted). As such, the lower levels of in-cylinder residuals being pushed back into the intake port with the Blair exhaust results in a intake higher charge purity and what charge is there is also at a lower temperature. The advantage of the latter is that lower temperatures result in a higher density, which in the case of a racing engine mean a higher mass of air and fuel being induced per cycle. The result of both these factors is that the Blair exhaust will allow a greater mass flow of air and fuel per cycle into the cylinder and as such will increase the engine torque/power further.
The resultant performance of the Blair exhaust is a 3bhp increase in engine power at 7250rpm over the std G50 exhaust.
It does however need to be noted that over the lower speed operating regime (below 6500rpm) pumping losses were actually higher with Blair's megaphone compared to the std G50. This would cause a drop in torque/power at speeds below 6500rpm. A few key conclusions about what this means on a track:
- The Blair exhaust system benefits significantly from a 5 or 6 speed gearbox. The 5/6 speed box results in lower engine rpm drops during gear changes and as such allows the engine to operate at longer engine speeds where the Blair exhaust has a benefit (6750rpm and above)
- Due to the more 'peaky' nature of the Blair exhaust, realistically only a top rider who can keep the engine spinning at 6750rpm and above would notice the improvement in engine performance. Quite simply, if more of a clubman/parade type rider kept engine speeds always below 6500rpm the Blair exhaust would actually result in poorer performance than the std G50 design.
The megaphone
produced and sold by Colin Seeley Racing Developments was slightly different to
Blair’s original design that was used on the Irish Racing Motorcycles G50 of
Mick Mooney and Ronnie Conn. Although slightly shorter in overall length, it
was still a very slow taper and without any form of reverse cone. This
race-proven exhaust became very popular in the early 70s and most Seeley G50
machines that were sold included this option. With it's unique sound and look, it was not long before it became one of the
most loved and distinguishing features of the Seeley machine.
Blair/Seeley G50 megaphone: Megaphone length 800mm
Megaphone
outlet 100mm
.
This comment has been removed by a blog administrator.
ReplyDeleteWe have one for sake if anybodies interested
ReplyDelete