MG-Cars.net

Welcome to our resource for MG Car Information.

Recommendations

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG MGB Technical - Go With The Flow, Or Get Left Behind...

A short while back there was a post called "flow question" that went all over the place, however I promised to answer a direct question and in fact, volunteered to post additional information regarding our procedures of cylinder head development in regards to airflow. If you have ever been curious, this would be a good post to take the time to read.

I initially planned only to give a brief description of the process and simply answer the question. The problem with that, is the more I got into it, the more questions I raised, until I decided to just go ahead describing the many aspects of airflow and the processes used to develop more efficient, "better breathing" parts. So let's begin with a primer in airflow...


Fire Breathing Engines?

The reciprocating internal combustion engine is often likened to an air pump and to a point this is true. However, the fact an engine runs at all is not so much due to any phenomenal pumping abilities, but rather the presence of the combustion event...Something we must keep foremost in our minds as we consider the following.

Why then does an engine require air?

Air is one component necessary to support combustion, fuel and spark being the others. The pressure excerted by the combusting gasses (fuel and air) is what pushes the piston and turns the crankshaft. It would thus stand to reason that anything we can do to increase the amount, or duration of pressure, will improve the performance and torque output of the engine.

By increasing the amount of charge (air and fuel mixture) inhaled by the engine at any given speed, we will find that torque output will increase accordingly. This comes about due to the increased volume of gasses acting on the same piston area in the same amount of time. The key here, is time and we will later find that increases of inhaled mixture may not produce the desired effects, if the burn rate is made to suffer as a consequence.

So what then is Ve?

Volumetric efficiency is the term given to the amount of charge inhaled by an engine of fixed displacement, verses the volume of it's cylinders, 100% being theoretically optimum. It is practical to note that highly tuned racing engines frequently exceed 100% Ve, and is the reason for their higher specific outputs.. However, considering the MG series of engines, and the 'B' in particular, one finds himself hard pressed to achieve 100%, even for full bore racing efforts.

Some factors affecting Ve include breathing efficiency and pressure wave tuning. Naturally one would assume that, were we to increase the efficiency through which the engine is able to breath, it will pump in more mixture with less effort. We will find that, while this is true, "breathing efficiency" is a multifaceted component and beyond the following statement, cannot be considered as a whole.

The other most notable factor, would be pressure wave tuning. Using the elastic and compressible qualities of air to our benefit, it is possible to increase the filling of an engine's cylinders up to and beyond 100% Ve. It is important to note that this is a fully developed subject in itself and will thus will not be considered in the remainder. It will also be noted as an unfortuante fact that siamese port cylinder heads do not lend themselves well to this form of tuning.

The next step..

For many years, engine builders have modified cylinder heads in an attempt to increase the performance of the engines they have built. Whatever the focus was on, (an increase in Ve, combustion quality, or both) the underlying goal was performance. For airflow directly, we will consider it's elements in a three dimensional context.

Dimension one, airflow Quantity:

Imagine two inlet manifolds for a single cylinder engine having one inlet port and carburetor (like a lawn-mower engine). In the first example, we will use a straight section of pipe 12" long and of a cross sectional size adequate for the desired output and operating range. In the second example, we will bend the above pipe in two places at 90º angles, effectively forming an acute dog-leg of constant cross section, this time sized to flow the same quantity as in our first example.

Given both units flow the same quantity of air, it may be surprising to find that in practice (on a running engine), the straight section out-performs the bent one. The reason for this, is that flow quantity, is only one aspect of total "breathing efficiency." It follows then that an engine must have sufficient quantity for it's desired output and operating range, anything more simply detracts from performance within this range.

Fortunately, the quantity needs of an engine are easiest to calculate. For example, an 1800cc engine running at 7,500 RPM will require 237.75CFM of air, per revolution. The forgoing example assumes we have obtained 100%Ve, which at that RPM, at least for an MGB, is pretty unlikely.

Dimension two, airflow Efficiency:

The fact is, a straight pipe will always prove to be more efficient than a curved one, regardless of flow quantity. This is due to losses encountered whenever air is made to turn en-route to a cylinder or other means of relative pressure depression. Because of this, the straight pipe can be made to flow considerably more air at a smaller cross sectional size than it's curved counterpart, and increased performance is the result.

Air flows due to a differential of pressure between two points. An example would be the air at atmospheric pressure flowing into a cylinder at sub atmospheric pressure. Additionally, airflow will always take the shortest route between these two points of pressure differential. I.E. airflow takes the path of least resistance. Therefor, air will "hug" the inside radius of any corner present in it's path from pressures high to low, while attempting to take the straightest route. This in turn causes unequal pressures in the port, with flow sheering and subsequent losses in efficiency the result. These effects could thus be termed as "resistance" to flow motion.

One way to decrease the "resistance" felt by the engine, is to increase the cross sectional size and thus volume of the manifold. While this does increase the flow quantity potential, it does not increase the flow efficiency, as turbulence remains much the same as before. We already know that smaller cross sections are desirable and we will now explore the reasons why.

Inertia tuning explained:

Inertia tuning or (ram charging) is a state of tune where-by the incoming gasses to a cylinder are provoked (by their inertia only) to continue flowing despite a lessening, or reversal, of the pressure differential between the atmosphere and cylinder. We learned in physics class that bodies in motion tend to stay in motion and while air is basically invisible to the eye, it none the less possesses significant inertia at speed. To understand how this relates to cross sectional sizing of ports and manifolds, let's carefully consider the following.

For a given pressure differential, the air must flow at a given velocity, to achieve a given output quantity, through a duct of a given cross sectional size. Therefor, if we decrease the cross section, we decrease the manifold volume and thus the velocity must increase to achieve the same output quantity. Since velocity increases inertia and therefor cylinder filling, one is compelled to size port and manifold cross sections to the smallest that will deliver the necessary volume at a known pressure differential. This results in the maximum cylinder filling with the least resistance to the engine, which is the essence of airflow efficiency.

Since cylinder heads and manifolds must fit into engine compartments as well as satisfy other practical requirements, most ports are not straight. Therefor the need to increase the efficiency of a curved manifold arises. Knowing how cross section affects velocity, it is possible to use these effects to increase airflow efficiency. Since straight sections can be sized smaller than a turn for the same output quantity, a port or manifold utilizing these characteristics will see an increase in efficiency over one of constant cross sectional area. Additionally, cross sectional shape can be used to enhance efficiency. By increasing the area along the inside radius of a turn, flow sheering and subsequent losses can be reduced. The most efficient manifold or porting arrangement, will likely utilize all of these design aspects.

Dimension three: Airflow Quality

When dealing with airflow in the internal combustion engine, it is easy to loose sight of the fact that intake air includes not only air from the outside atmosphere, but atomized droplets of fuel as well. This fact makes the routing of fuel and air particularly troublesome as degradation of mixture quality will have a directly imposing negative effect on the production of power. Remembering back to the beginning of the article, we will recall that the combustion process is critical to engine performance. The "quality," or state of mixture homogeneity, is one factor affecting the combustion process.

If the air/fuel ratio varies within the combustion chamber, (that is 12:1 in one cc, 16:1 in another and so on) the flame propagation will be impaired. When a situation like the above occurs, the flame front is forced to advance in a stop-start manner which increases the overall length of time necessary to complete the burn. The effects of this can lead to problems such as detonation and other losses, including low fuel efficiency. A simple way to see how mixture quality affects combustion, (and thus power and efficiency) is to change the air/fuel ratio metered by the carburetors. Owners of smog carbureted cars who have changed to non-smog needles will be the first to attest to the performance enhancements obtainable. In this case we changed the overall ratio only, however changes in airflow quality can have equally as pronounced effects. Additionally, this quality is directly affected by certain properties of airflow.

Keeping the fuel suspended:

Reasons for a non-homogenous mixture having instances of poor fuel atomization, are due to deficiencies in the airflow quality during it's route from the carburetor, to the cylinder. Since fuel is heavier than air, it tends to be "centrifuged" out of suspension whenever a turn is encountered in the flow. This problem can also be caused by the effects of viscous sheering in the air. If the port or manifold has been designed so as to minimize sheering of the airflow, atomized fuel will be less likely to depart company from the bulk of the flow, whenever a turn is encountered. Additionally, an increase in cross sectional area through a turn will slow the air and fuel mix as it makes the turn, thereby reducing the chance for the "centrifuging" problem mentioned previously.

Although cross sectional changes and flow velocity can be used to our advantage (as seen above), taken to the extreme, these effects can become negative. Fuel "drop out" as it is often referred to, is usually due to sudden cross sectional changes to the flow path. This has the effect of suddenly slowing the air through the change (in this case, from a smaller to larger cross sectional area) and the fuel literally "drops out" of suspension. Properly designed ports and manifolds will not have sudden or radical changes in cross sectional area. Any change needed must me made smoothly or the proceeding problems will result.

Let's finish on this:

Last on the list of factors affecting airflow quality, is port finish. The fact is, a properly finished port is critical to the performance of the engine, as it is the final word on airflow quality. However, when "port finish" and "high performance" are mentioned in the same sentence, most enthusiast immediately believe that the finer the finish, the better the airflow. This is an easy trap to fall into, as utilizing the "finger test method," one would consider a port which feels smooth, must flow "smooth" as well. Unfortunately, this is not necessarily the case at all.

The airflow through a port of an internal combustion engine is what is known by fluid dynamists as "fully developed, turbulent flow." This simply means that the flow is not of a laminar state, and instead consists of molecules of air continually flowing at different velocities all the way from the center of the port, out to the edge of the port. For a perfectly straight pipe (as in our first example), the molecules in the very center are flowing the fastest, with those closer to the edges slowing incrementally as we near it. The molecules directly in contact with the edge of the port are in fact, "stopped" for all practical purposes and this is all due to the viscosity of the air itself. Potential problems due to these effects include flow shearing, as more molecules attempt to go faster near ones which are going slower. This tends to help promote fuel drop out and centrifuging, and poses further resistance to flow motion.

Since we cannot change the viscosity of the air, we must instead work to find a way to keep the flow from adhering to the port walls. One way to do this, is to create a rough (textured) surface which in turn will promote what are called "Karman ring vortices." These vortices will "energize" the layer near the surface of the port wall, acting as some head modifiers have stated, "like needle roller bearings." The results of this phenomena act to reduce the adherence of the airflow to the port walls, allowing the bulk of the flow to proceed in a more homogenous, less disrupted fashion.

Just as cross sectional shape, size and consistency can be used to influence the airflow, port texturing and finish can likewise be used to our benefit. This means that, not only should a port be of a courser overall texture, but this texture should also be used to influence (or at least not degrade) the flow pattern in that specific area. If you haven't gotten all of this, don't worry, what it all boils down to is the fact that a finely polished port will not produce optimum results. We use anywhere from 36 to 80 grit on our manifolds and heads with the latter not being for the intake ports!


So knowing now how the various aspects of airflow influence the operation of the internal combustion engine, it is easy to see how concentration on one aspect only will generally not produce the intended results. For instance, concentrating on airflow quantities only, to the exclusion of efficiency and quality, may look good on a flowbench but in fact reduce output due to poor cylinder charging and/or fuel atomization, etc.

Next time we'll deal more with the development issues for a real eye-opener of how a flowbench can be used to reveal the true nature of airflow phenomenon.

Sean
Sean Brown

wow
do you have time to do any work
cameron gilmour

Thanks Sean, thats some heady stuff:)
Clark Burdick

Many thanks Sean

Very interesting - although I am unlikely to do any head waork myself, it will me help assess what is being offered if/when I spring for an upgraded unit. I'll print this out and wait for chapter two!

Regards,

BarryQ
73B
B.J. Quartermaine

Sean,

Thats a great article that really expains to the layman what design decisions there are around head work. Our thanks for taking the time to write it down.

Never one to be completely satisfied though, could someone explain in a similar manner how a cross flow head is able to generate such an increase in power and torque over the standard head? I understand that it is down to the different flow dynamics, but would be interested to know what engineering and design differences there are in the heads, and why these have the effect they do.

This could make for one of the most enlightening threads Ive ever seen.

Richard.
Richard

There's an article in the current MG World (went on sale on 8th Jan in UK) about an MGB that has been converted to a crossflow head.


Including all the other mods made (Dis, carbs etc.) it develops about 180bhp.
andrew.horrocks

Sean . . .
Great read. I thought I knew some but now I know more.
I read a lot of books etc and some of the better were written in the 50's or earlier. Comments like 'fuel injection might be the way to go in the future if it can be tamed and controlled'. I will look forward to more. I myself am electronics eng and am playing with a keypad adjustable fully programmable electronic ingnition system. Sofar it works great on the bench and the half in the car also but I get bored and it gathers dust till my next burst of enthusiasm.
Currently my head is being machined and exhaust valves/seats upgraded for ulp. Great board and top post.
. . . . .Henry
henryo

Thanks, Sean for a great bit of clear technical writing on a high interest subject. Not everyone who understands can make their understanding clear, and not everyone who can do the work knows where to begin to try to make what they are doing understandable. You should give thought to a book or such on this, if you haven't already. If you have a rare gift such as this, you ought to exploit it.
Bob Muenchausen

Richard,

The differences are based on the ability of a single port/cylinder arrangement to better fill each cylinder. In other words, there is no "cross-talk" between cylinders which impedes the inlet flow to them.

Additionally, the layout provides the opportunity for improved mixture composition (quality), which can be directly related to increases in power and economy.

I'll be looking to give the MSX a little recognition in the next installment. Please remember that (as Carroll Smith says), "I'm a full time racer and a part time writer." These things take time... ;)

Sean
Sean Brown

Andrew- That's Cameron's cross-flow car, he's the guy 3 posts above you here.Sean, speaking for myself, I'm ready to hear some good things about the MSX crossflow, besides small valves, warped faces, and slick packaging. I recall you saying you thought the MSX was the better potential compared to the Derrington.
vem myers

Vic,

My comments were based on the heads in the modified state and not as-received. I also said that I had no experience with the Derrington re-release, so the jury is still out on that one. I will say that my "new" post is about half done and focuses almost entirely on the MSX... Stay tuned folks.

Sean
Sean Brown

Just to keep this thread alive while we wait for Steve's next treatise, it seems to me, as one of the old school of try it and see, that lots of this theory doesn't actually help when it comes to knowing where to stick the grinder and how to waggle it about!

Much of the discussion has centred around how to remove material from the ports to enhance airflow and thus power.

I just wonder whether a different approach might be worth exploring ( or indeed if anyone has tried this in the past) - how about controlling the airflow by inserting material into the ports?

My aero engineering background suggests to me that it may be possible to insert aerodynamic vanes into the ports to increase airspeed and direct the flow more precisely into the valve "curtain" area, reducing the cotton wooliness of the air and turning it more into spaghetti.

Perhaps if Peter would sell/lend/hire one of his engine dynos to me I could set up the test mule and compare the various offerings and test my theories - cor, I could even write a book on it!
Chris Betson

Funny you should mention this - there's an infommercial on Speedvision (TV) here in Canada that shows a thin metal spinner that fits tightly in the hose between the filter and the intake plenum. The idea is that it spins freely in the airstream to introduce turbulence and increase velocity. Of course according to the retailer it dramatically increases performance, economy and smoothness - don't they all :). Is there any merit to their claims?
Mike Polan

I really can't imagine that there can be much if any benefit from a free spinning vane - it was not this sort of gimmick I was considering but rather a number of longtiudinal guides that helped the airflow follow the contours of the port and turned it towards the valve head so that flow was more controlled and, because the free port cross section would be reduced, speeded up - aiding low rpm torque.

Steve claimed in the early thread that the laminar flow detached from the port wall where it opened out past the pushrod tubes and it then headed straight across the port towards the central dividing web.

My thoughts centre around introducing a tapered cone like device attached to the web and shaped to deflect this detached flow towards the open valve head.

The last thing we want to do (from my experience with Concorde engines) is to induce any "spin" to the flow - just like water going down the plughole, this will slow the entry of the mixture into the cylinder potentially by a factor of 2 to 1.

Ideally the flow should run axially down the valve and flow evenly through the valve curtain, but this means turning the flow through 90 degrees with as little loss as possible - which is what I reckon what Peter's skill ( well Shaun's actually!) is all about....and why xflow hemi heads are better at because they only have to turn the flow thru 60 degrees or thereabouts.


Chris Betson

Sean-
You said "there is no "cross-talk" between cylinders which impedes the inlet flow to them." This is a headache common to al siamesed-port designs. Could you explain the dynamics of the "cross-talk"? Just exactly what happens, in what order, and what can be done to reduce it without going over to an independent-port design, such as a crossflow head?
Steve S.

Sean- Yur gonna get questioned to death here, better to rush that MSX treatise to early press.And when you come back , tap your headlights, so's that we can roll out the red color carpet
vem myers

Chris,

Vanes and such won't help because they assume a steady state flow condition. In reality the flow will constantly be changing, so they would get in the way more than help. They also should not be needed were the ports designed better in the first place. Port work always involves compromises, but the 5-port heads are the biggest compromise of all, there is only so much that can be done in the space available.

Mike,

Those are just as bogus as the ones you place in your exhaust pipe for the same reason!

Steve,

I don't think anyone can fully answer your question about the dynamics. I can say that there are times when the flow will go into both cylinders, and times when it will go into one and come out of the other. What can be done about this, is you can pay attention to controlling the "cross flow" from one cylinder to the next. This can be done not only through attention to port and valve/seat modifications, but also in camshaft design. You are undoubtedly aware of the scatter-pattern cams on the market. I feel these are a good step toward optimizing the power available. I would like to see cams such as these available for the street market as well, but development money seems to be tight for those concerned.

Vic,

About one or two days and it'll be "press time".

Sean
Sean Brown

Sean-
I don't seriously expect a full answer my question about the dynamics. A partial explanation, much like your original posting, would do well enough. Just how do you go about "controlling the "cross flow" from one cylinder to the next"? It's a fascinating subject, as we all know.
Steve S.

Steve,

This is indeed a complex subject, where the more you know, the more you realize how big a problem you are dealing with. In general, I will say that the more "oriented" and "homogenous" the flow, the less it will tend to reverse course. That is, the more efficient you can make the port (where ALL the air is going in the same direction with a reasonable pressure distribution throughout the cross section), the less likely you will be to have "pockets" of "stagnant" air that will easily reverse course, or cause the bulk of the flow to more easily do so.

How this relates to the siamese port heads, is that the more efficiently the flow goes into the first cylinder, the more "hesitant" it is to swap over at the instigation of the proceeding induction of the adjacent cylinder.

There are ways to do this with the addition of material to the ports, however the volume of the port also plays a role in the picture, the larger the siamese section, the more the size tends to "dampen" the pulses and subsequent flow. This is something we have found difficult (at best) to predict and more so to tune.

Additionally, attention paid to reverse flow will help one design seat and port details that will discourage reverse flow, and subsequently flow from one cylinder to the next.

We also pay attention to the form of the divider and make sure our porting in this region takes into account the want of the air to swap from one port to the other.

Lastly, you can plot piston motion and valve flow capabilities on graph paper (as we have done) to get a very serious visual representation of how camshaft timing events can interact with adjacent cylinder displacements and piston velocities. If you do, you will find that with even somewhat "mild" cams, the situation is not good! Doing this also allows one to find the points of valve lift that are most "critical" to the development of reverse and forward flow.

I hope that answers your question. I think for an engine builder or your experience, you can appreciate how first hand knowledge of these things can lead to a better understanding than knowledge acquired from that which is related.

Sean
Sean Brown

Sean-
I believe that the pressure waves in the intake port resulting from the 180 degree throw difference in the crankshaft has a definite influence on fuel/air mixture separation and fuel condensation in the arriving fuel/air charge and that this is what creates the impression that the inner cylinders run rich. I believe that the rich mixture on the inner cylinders is in reality the consequence of the problem of the interplay between the resulting stuttering flame propagation that you mentioned and reduced atomization of the gasoline caused by the return pressure wave. Do you agree?
Steve S.

Steve,

Really interesting point, I'm inclined to agree, the colour striations in the carbon deposited in the chambers indicate interupted flame propagation in cyls 2&3 - almost like ripples on the beach - 1&4 are much more evenly coloured and grade out from plug to wall.

Sean,

Do you think that the fact that cylinders 2&3 have siamesed exhaust ports whereas 1 & 4 have single exhaust ports leads to greater or less scavenging in the centre cylinders?

This must have some effect, especially with the wider overlap cam profiles.
Chris Betson

Sean-And it continues. I'm holding my breath to press time. Steve- The simple remedy is go to cross flow.
vem myers

Vic-
As you know from your own experience, only in theory is going with a crossflow head the simple way to solve the problem. Buying the head and all of its associated hardware, plus new intake manifolds, a pair of Weber DCOE carburetors and their associated linkage system, new airfilters, trick headers and exhaust system, etc- it all adds up to a very tidy sum indeed ($$$$$!). Then comes the fun of choosing a camshaft that's appropriate to your intended use for the engine and paying for time on a dynometer so that the jetting will be right! Small wonder that most people prefer a reworking of the head that came on the engine!

Sean-
What do you think of the scavenging issue that Chris brought up? And what are your thoughts on my pressure wave idea?
Steve S.

Steve,

There is no doubt that the adjoining cylinder's "overlap" with the first's makes for a poorer mixture quality. Obviously, auto makers have not gone to single inlet port per cylinder arrangements because it's a step backwards. However, there are many things which degrade the mixture on it's route to either cylinder in a 5-port head. Even if the valve events were not in sequence the way they are, the mixture would still not be as homogenous as were the ports of a more efficient, single cylinder/port design.

Chris,

Regarding the carbon tracking: One thing you will notice with any 5-port head, is a "clean spot" on the chamber walls around the inlet valve. This is where fuel is/has separated from the mix and literally "washed down" the walls on the inlet stroke. lack of carbon denotes lack of burning, and it's pretty obvious that these areas are not burning well. You will usually find evidence of overly rich combustion in the region of these clean spots also. This is why attention to flow quality is important. Reducing these "clean spots" and getting the fuel more atomized is the name of the game if more power and economy are to be had for a given level of airflow.

Regarding the exhaust ports, the end one's tend to flow better. So given the right exhaust system, they should scavenge better. But it would depend on the header/system in concert with them. Remember, the adjoining exhaust ports are not in "sequence" like the inlet ports are.

Sean
Sean Brown

This thread was discussed between 09/01/2003 and 21/01/2003

MG MGB Technical index

This thread is from the archive. The Live MG MGB Technical BBS is active now.