RZ350R Porting: Worked Example

Introduction

This document provides a worked description to using the software provided in the books by Professor Gordon Blair, available Here. Theis software packages may also be available on the internet as the first book is quite old. These can be found easily using a search engine.

A worked example is provided for my old Canadian Yamaha RZ350, 1985. The procedures used to evaluate this engine can be applied to the engine you wish to evaluate.

Substantial data collection is required to evaluate an engine, more if a simulation is to be performed. You will need lots of patience, and be able to do high school math using a scientific calculator. A large handful of patience and careful methodology is required too.

Data Collection

The first thing that must be checked is the engine condition. No performance modifications should be attempted on an engine that is not serviceable. Bore wear, bearing clearance should all be within limits.

Bore rubbing

A piece of paper, rolled into a tube and taped into the barrel needs to be cut, the length is appropriate for the barrel length, and the width should be approximately 3.2 times the bore diameter. Once taped in place, a wax crayon or soft pencil can be used to make a rubbing of the port outlines and edges. Note that, in some engines as found on the RZ, the exhaust port is most generously radiused. The actual port area will need marking on the rubbing.

Also, the carburettor diameter will need measuring, as will the reed cage assembly. This will be discussed later.

Head Volume

This will also need measuring - take a flat alloy plate large enough to cover the gasket face of the head and drill a small hole in the centre. Place this plate onto the cylinder head, applying some grease to the gasket surface. This grease will form a convenient temporary seal. Fill the head via the hole using a calibrated measuring device. This is often a burette in available literature, but these are expensive, and a 0.5 cc-graduated syringe from a Drug store is suitable. This should now give the head volume in cc. With the engine assembled, and the barrel bolted down but without the head fitted (this may need spacers under the cylinder head studs if these hold the barrel in position too), measure the distance the edge of the piston is from the top edge of the barrel at TDC. This is the so-called deck height. Also determine the position of the piston with respect to the ports at bottom dead centre since the ports often descend below the piston edge. Determine the thickness of the cylinder head gasket as fitted - usually in its crushed condition. Add the deck height to the head gasket thickness, and determine how much volume the piston deck height takes up by multiplying as so:

Finally for the head, the squish clearance needs measuring. Using 4 lengths of plumbers solder, these must be stuck to the piston using grease in a north, east, south and west configuratio, with the centre of the piston the centre of the directions. The head needs fitting and torquing into place, and the engine turned over once (using a spanner on the flywheel or main gear/clutch drive). This crushes the solder, and when removed from the engine, the head/piston clearance can be measured. These should be very similar with a centrally located spark plug arrangement as used in the RZ. Notice it is sometimes necessary to skim 5 to 10 thousanths from the piston crown to correct for non-central piston crowns. This is also useful since it allows parallel combustion chamber design since the exact piston crown angle is known. Any port timing corrections can be made by skimming the base or top of the cylinder as required - this is often necessary anyway to obtain the correct deck height necessary for squish clearance, and also to ensure correct opening timings of the ports.

Port Timings

The rubbing taken above needs measuring and turning into values suitable for the next stage of evaluation with the computer. Once the paper is opened out, the widths of the ports can be measured. However, these are arc widths, while we are interested in chordal widths. Therefore, the widths need correcting using the following formulae:-

Now the port duration can be calculated. Either use Prog11.exe, which gives a tabular form suitable for printing out and pinning to the workshop wall, or my handy on-line utility, or even Degreeit.exe by Tom Turner if you have it. All these utility programs require the stroke and connecting rod length of the engine being worked on. REMEMBER to correct the actual port heights for effective port heights (relative to the edge of the piston) since often the piston does not fully uncover the ports. In this situation, we are presuming the exhaust port is fully open at BDC.

Construct a table (similar to the one below) of parameters, possibly using a spreadsheet.

The inlet port is listed, but not for duration. This is because the inlet uses reed valves to control duration; the port area is of interest here when dealing with matching areas later.

Time Area Analysis

Time area is a concept used to normalize different ports against different engines. The time area is given in mm squared per second per cc, in other words port area open per second per cc. There are empirically derived guidelines available for this, but these are just guidelines. The time areas are not the sole ingredient - carburation, expansion chamber, ignition, head etc. all have a vast influence on the power delivery of the engine, and as such must be matched to the chosen operational speed. An engines power band is the distance between the peak torque speed (usually lower in the rev range) to its peak power speed. A race, high BMEP engine will have them nearly identical whereas a road engine would have say 3000 rpm difference between them.

Obtain the desired time area targets for your chosen engine, either using a time area target calculator such as target.exe or prog61.exe

Alternatively, analyze an engine you would like your engine to mimic. Good examples for road machines are enduro engines - road race engines and motocross are too highly tuned to be reliable.

The time area target programs ask for a desired power output at a given engine speed. We choose 10500 rpm and 40 bhp from a 173.5 cc engine (since the RZ350 is twice this, and we want to make 80 bhp); there are approximately 0.75 kWatts to a Bhp. This gives

Notice there are two values for the time area targets. This is governed by engines with small ports and high crankcase compression rations for small time area, and larger ports with lower crankcase compression rations for the larger time area. Experience is the only help in choosing which values to employ...

Also notice the time areas are total time areas - this means the sum of any individual port types.

Now we must calculate the time areas for the engine under test. Prog63.exe is suitable for this. Enter new data following the menu scheme. 7 mm and 10 mm was used for exhaust port upper and lower radii. Each of the transfer ports must be calculated, and the time areas summed to get the total time area. The program asks if a bridged exhaust port is used. A table of ports and data can then be constructed.

The time areas are given x10^4, so the figures need dividing by 10000 to get corresponding values, since the targets are given in seconds per metre whereas the values here are seconds per mm. The only blowdown we are interested in is to the port which opens immediately after the exhaust: in this case, the Boost port.

Giving:-

This means the exhaust port does not have enough time area for the speed we want. There is too much blowdown, meaning the exhaust opens a little early. The total time area for the transfers is within the accepted range too, but towards the lower end.

If the exhaust port is generously widened to 45 mm and 12 mm radii port windows are used, then the time area is more like what is expected. Modern engines can easily tolerate 65% port width on ductile iron rings. Generously radiussed ports can have up to 70% port width, going up to a maximum of approximately 72% for a racing engine whose rings are frequently changed, with a very oval port. The shape of the port is also important. Rectangular port produce very sharp pressure pulses, which can make an engine either peaky or make the exhaust work better. They definitely make more of a high pitched noise which is unpleasant to people though. An oval port produces pulses that are gentler and easier to silence, but can make the exhaust not as useful as it might be.

Raising the port is generally the last resort for area, since it reduces the effective power stroke of the piston. It also raises the peak torque limit too.

Modifying transfer ports are usually quite difficult to perform properly. A lot of home tuners simply file the windows directly, but this is incorrect since the nozzeling of the port is not changed. Often, the transfer ports extend below the piston edge, and more advanced timings can be created using a barrel spacer. This also has the disadvantage of raising the exhaust port too, and also the direction of the ports remains the same. It is up to you to determine if this modification is worth it.

Comments

The port timing and time area is not the only concern when evaluating cylinder barrels. Cylinder and liner alignment needs to be checked and corrected where necessary. The wall of passages needs grinding to a satin finish - mirror polishing is usually not required, but some tuners like to create this finish in the exhaust port so carbon deposition is kept to a minimum. Another common modification is to thin the central wall in transfers, creating a so called "knife edge" effect. This is a useful modification when not taken to extremes, since the flow should be improved. The whole idea is to create a generously tapered nozzle, encouraging the mixture to increase in velocity but not break up and cause turbulence. If this separator is too thin, then the barrel strength is weakened. Another common modification is thinning central bars in ports. I am against this personally since this puts more stress on the piston skirt, which could break unless a race engine is the target application. The RZ currently evaluating has a terrible step on the bottom edge of the transfer port that will require filling with metal loaded epoxy and grinding flush.

The exhaust port has an area that should match the area of the exhaust flange that fastens the expansion chamber to the barrel. Sometimes the exhaust downpipe must be enlarged to keep flow optimal and match cross sectional area between exhaust port, exhaust flange and exhaust downpipe.

Head Reprofiling

Once the barrel is designed, the cylinder head can be evaluated. With the head, we are trying to achieve a desired compression ratio, a desired squish clearance and desired maximum squish velocity. There are a few pointers to MSV on the WWW, and you would be advised to read them. Correct MSV ensures optimum mixing at the target engine speed. The same goes for squish clearance, since this effect helps prevent detonation. The compression ratio helps obtain the maximum energy from the mixture.

Generally, trapped compression ratio is around 9:1 on pump gas, squish clearance varies between 1 mm and 0.5 mm, smaller for little engines, less for larger engines, and MSV is 18 to 22 m/s.

Unfortunately, unless you own the comprehensive software set by Tom Turner, it is a matter of employing programs Prog42.exe and Prog41.exe until the numbers come out right. Prog42.exe allows you to enter the current status of your head, and Prog41.exe checks it for the MSV. Once the numbers are to your liking, the head and barrel must be modified to reflect the numbers used in the simulation. This usually means skimming the top of the barrel to adjust the deck height and reprofiling the cylinder head.

With the RZ engine under test, the Area ratio = 0.3, Chamber radius = 41, Squish and piston radius = 152, clearance =0.75 gives 10 cc clearance volume and 9:1 compression ratio. This gives MSV of 18.8 m/s at 10500 rpm.

Reed Cage Analysis

Prog64.exe allows analysis of the reed cage. It predicts a carburettor size for given reed block dimension. The user must enter number of petals, petal size and material along with cage dimensions. The effects of modifying the cage can also be investigated, for example the fitting of Boyesen one piece reeds allows the central bar to be thinned, giving increased flow. The program also calculated the optimum bending shape for the reed stopper plate. Unfortunately, due to time constraints, this was not performed.

Expansion Chamber design

Once the barrel, head and reed cage has been evaluated, then it is time to examine the expansion chamber. This has the most pronounced effect on power characteristics. If a chamber is designed to resonate at the target peak power frequency, then the engine will be very peaky, but very strong on that power. The user has the option of designing a chamber on the power, on the torque before using the engine porting to create over rev or design the chamber to over rev. There are 2 versions of Prog62.exe : one produces expansion chambers with a 2 stage diffuser and is DOS based, the other produces expansion chambers with a 3 stage diffuser, and is Windows based. Examination of current racing machinery makes one think the larger the number of diffusers then the better control is given over the power band, but become more difficult to make. Alternatively, use my on line utility which calculates exhaust temperature based on BMEP.

The down pipe diameter needs making the same cross sectional area as the exhaust port window, so calculate it using the area given by the Time area calculator. The mid section diameter should be 3 to 3.5 times the diameter of the down pipe, the larger the better but more difficult to fit on the bike. The exhaust temperature depends on the BMEP of the engine, and there are documents available on the internet which discusses this. The exhaust type depends on the width of the power band, and is self explanatory. The horn exponent affects the shape of the chamber and 1.5 is a good starting value, and is usually between 1 and 2. Both versions of Prog62.exe are designed to produce expansion chambers using the rolled cone method of construction, and can produce marking instructions for cutting flat plate sections ready to roll.

Conclusion

A brief worked example of using the programs available from Tuners 2-stroke page has been given. These should put you in the ballpark for the engine power delivery you want. However, the only way to really develop two stroke engines is to spend dyno time with them, making incremental changes using several sets of cylinders.

Written 5^thApr 1999
Updated 18^thSep 2007