Wednesday, June 25, 2008

Tuned airbox design considerations for a modern day car. Tuning the airbox - a two stage idea and computer simulation model. - unfinished yet

This is a hard one. First some links with some math on how to get the resonant frequency of the airbox. What is an airbox? We call the intake pipes with the distributing channels an "airbox". Its purpose is to provide more air with less resistance due to the pressure oscillations driven by engine air suction. Properly designed air intake system can provide 20%-30% power increase over no intake system at all.
It is important to make all parts of the intake well designed and properly tuned to the DESIRED RPM range. 2-cycle engines are the most sensitive to good or bad changes in engine responsiveness and performance to intake system change. (exhaust too, as the petrol-air mix travels to the exhaust, but is then reflected back.)

Also, calculations allow us to understand at which frequency was the original intake system designed, so we may be able to make a completely different hardware design with the same frequency of resonance or at least the same characteristics. In case change the resonant frequency we may be able to estimate how much did we deviate from the original air intake.

Short and easy to understand text written by a motorbike racer, summing the experience and basics behind the operation: (a good starting point text)

A very good scientific approach on the topic, using inertia, speed and conservation of energy rules to make the frequency calculation [WARNING units there are NOT consistent, nor are in SI, prease rewrite those constants for your own use, also careful - perharps 2-cycle engines discussed there??]

Next example considers the effect of "cold" air intakes and that they can provide ONLY 3% power increase at 150 MPH on a motorbike, so that the resonant airbox is the main source of power increase.

A simple and easy entry in the best physics encyclopedia on the net.

Some empirical calculation of an Helmholtz resonator.

In this moto thread the main topic is calculation why the stock air filter may not be that bad idea and some more.

For anybody interested in some practical examples of intake tuning on sports cars(including 2-stage airboxes)

Tuned airbox design considerations for a modern day engine.

And now my own research compilation and writing.

Firtst part deals with explanation of the parts used in the air intake system and manifold, next describes three major groups of tuned air intakes. First (A) consists of long straight pipes connecting engine air intake ports and a common air source. Rather 1960's solution. Next are (B) the Helmholtz-type resonant airboxes that apart of consisting of shorter lenght pipes connecting engine intake ports with a larger volume "airbox". Then we can see even better air intakes, of very complex design, CFD modelled, with many resonance parts, acoustic with superimposed flow, thermoacoustic design. Parts (D) and (E) display some of the well designed air intake systems.


Introduction to electric basics

In electric circuits we have inductors and capacitors. Wirewound inductor does not transfer lot of electric current as you apply voltage to it. It takes time for the current to build up. But when you suddenly interrupt it a spark occurs. You can store energy in an inductor but you can't hold it for long. Next you have a capacitor. Capacitor is very easy to fill up with electric charge and can hold it for long time. But as the capacitor gets charged, the current flowing into it decreases.

With these two components alone you are able to construct a large variety or circuits. The first is LC circuit, the two components in parallel will form a good resonant circuit. Whether you apply or remove current from the circuit an oscillation will occur as the two components will exchange electric charge at the resonant frequency of the circuit. If not periodically fed with changing current supply, the oscillation will gradually cease.

With different LC components connections you will be able not only to make resonant circuits but also filters, low pass and high pass. Low pass allows direct current and slow current changes to be transferred trough the filter while high frequency waves are blocked and high pass filter allows only waves of higher than some frequency to pass while direct current has no effect on the other side.

Compressible fluid mechanics

Aerodynamic equivalents of the above electric components are inertance and compliance. The inductive element, inertance can be made from a narrow tube where gas has to move with high velocity and thus carries energy in a form of quickly moving mass of fluid. The capacitive element is compliance. It is a vessel with large volume where we can store large amounts of air without moving contents of the compliance volume much. The energy can be best stored in a form of pressure increase.

Electric and fluid model differences

Every capacitor has a parasitic inductance. Also, every inductor has a parasitic capacitance. But in power components the differences of the parasitic inductance of capacitor versus the inductance of the inductor is like milion to one. (1000000:1). The same is when we talk about capacitance.
In fluid model you see that the inertance tube has also some volume and that the compliance width is not infinite, common differences are in the range of about 10:1. Also in fluids you can have travelling waves both in compliance and in inertance, but we use those terms to be specific about our intention of what the component will do. So you see that in fluids design it is rather easy to use a single component to achieve some repeatable results in terms of self-resonance. In electric designs we do not usually rely on how much the parasitic value is.
Next large difference is that in simple low-frequency electric components you only have current and voltage to deal with. In fluid dynamics, you not only have a flow with a certain density, flow speed, temperature and pressure. Above that you have also sound speed at which sound waves of the oscillation will propagate. And guess what: all these 5 parameters are interrelated and affect each other! So a perfectly designed air intake system either requires excessive physics knowledce and endless ability to simulate the design on a computer model, or good knowledge and lots of trial and error experimentation or both. The last method is used the most. "Trust, but verify."

What is this all about is to make passive components of air intake system that help the engine to get more air for the internal combustion. We need the air to burn the fuel with and to add to the expanding hot gases. The next part describes the practical implications of the above.

Air intake system designs

A) Inertance only tuned intake
B) Inertance-compliance tuned intake
C) Advanced and multituned inertance-compliance intake
D) example: Honda intake resonator, engine type D16Y7
E) example: Honda intake resonator, D16Y5 and D16Y8 engine types
F) ??? to be written
L) Where did we get the power to compress the air, again?

A) Inertance only tuned intake
Similar to the single straight cut exhaust pipe, typical for many Mercedes and Chrysler cars, the design of the rest of the air intake system is either separate or completely omitted. A short explanation here.

It uses the self-resonance and inertance of an intake manifold pipe, time delay and spring modeled design. Chrysler had it designed to peak at 2800RPM at sixth harmonic. You can surely imagine what the efficiency of a spring-resonant system working on 6th harmonic is. Example of realisation of the 77cm intake in the Chysler on the right.
[Explanation: imagine a pendulum of a clock. It swings to left and right in one second. So you swing it and on sixth swing you use the energy still remaining in the pendulum in its peak height. Not a problem for a clock pendulum, but if you swing a garden hammock you will notice the swing amplitude decays quickly. Plus part of the pressure pulse does exit the pipe and is not recovered back. Next design step takes care of that.]

B) Inertance-compliance tuned intake
This is a smarter design in the efficiency, no need to resort to 6th harmonic [ed. note: eww], it works in a similar way: the flow of the air that is flowing to the engine is abruptly stopped, so as a result the flow ceases and the energy of the inertia is converted to a pressure increase. What do we do now? We obediently return the resulting wave ripple back to the sender. But as it exits the relatively narrow inertance pipe it discovers wide and open space where its energy can be safely stored with little losses. So this wave travels trough the compliance volume to the other side where it either bouces back or finds the other inertance tube where it compresses the air. In this mode it acts much like an Helmholtz resonator. All four-cylinder motorbikes use this type af air intake. The good point about the design is that you can use simple shapes (square shapes of a box and single diameter pipes) to get good results. Also, the design and calculations are well predictable.

Remember that while some people anxiously try to minimise losses and save amazing hundreds of pascals (say decreased the pressure loss by 400Pa) the resonant system is all about making the pressure to go up and down periodically until it achieves amplitude up to 5-10% or more. 5% amplitude means 5% of the surrounding air pressure, that is 101.325kPa; 5% is then 5000Pa over and under surrounding air pressure in the peaks. So you see that our design goal is not to minimise the pressure loss, since the gains are small, but making the amplitude high and in correct phase of the air flow with suction in the engine intake ports. Yes, it needs to provide the pressure at the correct time AT the intake port, what pressure and when they occur elsewhere in the air intake system we don't care that much. Conclusion: you can reduce whirl losses later, once you have a working resonator.

C) Advanced and multituned inertance-compliance intake

We can do more than tuning the intake manifold only. It would be really naive to think that once you have intake manifold with good resonance that the resonating pulses will just stay there. In fact, you can observe at some engines that the air is being expelled out of the intake itself! So the pressure wave we so hard tried to build cheaply escaped from being useful to us. We can do two things to correct that. One is to make sure the pressure wave has no easy way of escaping the compliance fox of the manifold (the airbox), the other is to use the escaping pressure wave to promote further resonance of the air intake system as whole. And as you already can imagine, it is not possible, nor meaningful to completely containg the resonant pressure in the airbox, so both principles will be used. Parts (D) and (E) show one such system where the air intake system as a whole is about the same volume as the engine itself!

On a side note: Honda car engines typically perform at 100HP/1L of displacement as stock sold.

D) example: Honda intake resonator, engine type D16Y7

It doesn't surprise that to make a small engine with high power you need to be efficient and effective and that the Honda design again is. If you look at the size of the air intake components you will see that compared to the engine size, it is enormous. By "Effective" I mean "it does work" and also that it has good operational principles. I would like to point your attention to the fact that the pipes there do have their endings abrupt or rounded exactly according to the purpose they have! Some people would like to have round edges and large diameters on everything. Wrong. Even the small pipe leading to the front compliance volume with straight cut ending has its purpose. It forms a separate Helmholtz resonator, plus the separator in the middle also makes the resonance properties better (I think it is there to extend the resonant frequency). The pipe ending that leads to the air filter also serves as a gas diode, it allows the gas flow go one way easily and the other way is much more restricted. If you count with me you see 4 inertances, 2 compliances, and 1 double compliance box and 2 turns. Out of this the intake manifold is only 1 inertance and 1 compliance. See the difference of this complex air intake design and the simple long tube design from the 1960's Chrysler?

E) example: Honda intake resonator, D16Y5 and D16Y8 engine types

This intake system has a simpler compliance/helmholtz resonator on the air intake, but you can see something interesting in the straight piping right after the air filter in the schematic diagram. It looks much like a load cell to me. If it is a load cell it may be there to limit the maximum amplitude of the resonance; That alone leads to the decrease of the Q factor which in turn widens the resonance frequency peak. It may also serve as a separate resonator to buffer the sound pressure wave returning back to the air filter to protect it from high frequency vibration.

F) Optimal resonator, Optimal piston suction diagram.
2-piston compressor, wave pattern

G) Choice of resonance designs depending on engine design.

From a technical design standpoint we can put resonator designs into two categories: pipe-to-wall resonator and pipe-to-pipe resonator. It is not exact name, but you understand it when you see one.

Pipe-to-pipe resonator inertance tubes are not in the same phase of the resonant box. The opposed pistons compressor air resonator is a typical pipe-to-pipe resonator, pulse from one inertance travels trough compliance to the other inertance. This design however can only be applied in systems where it is physically possible, it is also not applicable to 3-cylinder engines too! Advantage: the oscillation is driven two times per one cycle, and each pressure pulse fed into half of the total number of inertances. As you can calculate the volume of the compliance and inertances of your design you will see that it makes a big difference.

Pipe-to-wall resonator inertance tubes are in-phase. Engines with suction ports in regular intervals are typical candidates for this type of design, also space constraints many times do not allow to apply other designs (motorbikes, cars). Disadvantage: since all inertance tubes are in the same phase, they load the compliance resonance more, also pulse is distributed among all at the same time, but only one of them has advantage of it, and per resonant cycle only one inertance tube can promote the resonance.

Volumetric efficiency?

I prefer only measuring how much mass of fresh air we managed to entrap in the compression volume. Consider that the whole volume enclosed by piston, cylinder liner and head is MORE than the displacement volume. So, should we count 100% VE as mass of air that would will the displacement volume and the volume in head cavity, piston cavity and prechamber at 101.325kPa and 23°C? Plus, if we manage to get inside some frozen air in cubes and double the total air mass entering combustion, would it be 200% volume efficiency? Therefore, lets understand that we need and want to know the total mass (weight) of fresh air we manage to use for the combustion process. Also, "volumetric efficiency" of 90% and 110% can mean the same amount of air molecules, depending on the conditions.

H) Stock Mercedes example, Suction pattern, loss analysis

hypothetical airbox1
- the good and bad points of this design become obvious
Good points: the low height compliance volume works more like a waveguide and indeed we can predict wave propagation inside of it in 2D simulation well. We can also see how a sound wave propagation would go if we use a vessel of the same shape and fill with water. A droplet of water in one corner will show how a sound wave will travel across such a rectangular box. That is a great advantage of these 2.5D manifolds.

Minimising pressure runaway back to the air intake and air filter
Hypothetical airbox2, improvement obvious, resonance
Improvement 2

Straight inertance to intake port with rounded edges: rounding edges causes little widening frequency response, small change in the base frequency, less static air drag, slightly decreased impedance and tiny bit increased parasitic volume of the inertance. Suzuki uses this system as their manifold is of constant internal diameter and maximum amplitude of the resonance is very important - motorbikes use mostly high RPM operation in narrow range.

Horn-shaped inertance to intake port: tapering the inertance tube so that the diameter is narrowed or widened at an angle. This causes things to be complicated. While it is predictable what it will do with static air flow, with pulsed flow it has the following effects: increase of the parasiti volume in the inertance, better potential to compress air at the narrower end, but it easily escapes due to the same reason, the resonance is much wider due to the interference and reflection from the steeped wall surface and also of lower amplitude compared to simple straight pipes. Despite the ugly disadvantages when inproperly applied, they have their place, ideally as straight air intake where each intake port has its own "horn" to surrounding air without filter. Per unit of intake manifold weight this is the best solution.

Relationship between inertance and compliance in an intake manifold
From the above you see that sometimes the design leads you to a point where you create a manifold which has long, wide pipes connected in parallel, equal distances to a box of small volume.
Not a problem by itself but consider this: it 1) decreases the maximum amplitude of resonance you are able to achieve, 2) even the pulse you manage to make you equally distribute among say all four pipes, thus minimising the resonance benefits and 3) all changes in diameter cause energy dissipation by converting one form of stored energy to another.

Flow rules.

a) If you have a flow of fluid trough narrowing passage, two things can happen: either it passes at the end with increased velocity at the same pressure as before, or the pressure at the end increases while the flow speed remains. [Very simplified, usually part of both happens and also temperature changes too!]
b) Plus, if you manage to increase the pressure by suddenly obstructing a flow of a fluid, a sound wave of this event will get reflected back and the flow will slowly revert due to the obstruction and due to the fact of the higher pressure. You may have noticed in your house that when you suddenly stop large flow of water, the pipes will make a "tikk" sound and also that the overpressure protective valve will let some water go.
c) With air, we are interested in these parameters of air in front of the intake valve: pressure (density), speed, direction of movement. We want all three to have right at the same time. I hope you already understood why and what can go wrong.
d) Conversion of energy forms is less than 100% efficient. Having horn-shaped inertance pipes may help air compression and low flow resistance, but at the same time it impairs resonance amplitude and introduces new losses from multiple flow conversions. Therefore, use with geat care and consideration!
e) with the point d) clear you have to ask yourself: "do I want to make a wave or do I want to make a flow?" - ask yourself honestly: "what is the mean flow speed and what resonance amplitude am I going to see?" As the flow speed increases the resonance makes much more sense, but at the same time the aerodynamic flow restrictions increase too. Don't overdo one and don't forget the other!

dalej: opisat priklad mercedesu, jednoducheho Mercedes Helmholtza, vylepsit ho o rozsirenie
uviest priklad sania ktore by sme chceli pre idealny rezonator (dvojvalec), sinus, a realny stav stvorvalca opisat rozdiel v umiestneni sacieho potrubia ako daleko od seba -> protilahle potrubie vs. protilahla stena -> rezonancia v polvlnnom mode a v celovlnnom mode
Realny saci diagram a rezonancia na n-ty piest. (stale hovorime o zakladnej frekvencii!)

It is a design that tries to build high pressure, but fails to contain it as it easily escapes to the air filter as the compliance diameter is about the same as the air inlet diameter.

Notice that we did NOT change the frequency of the resonant box, neither did we desperadely use n-th harmonic frequency, we still operate the resonator at its efficient, fundamental frequency and the only thing that changes is the piston that catches the next resonant wave. Isn't life great? Thus the problem we originally had with the resonant frequency selection solved itself. Well, at least partially.

* tuning intake with a turbo? *turbo and supercharger separation *overlapping valves and turbo
*invention-turbo with external valve *zmena casovania so zmenou plniaceho tlaku

[B]As for the air intake:[/B] to sip 600ccm of air trough, say 8 square centimeters area in 10 miliseconds, you need the air to be moving at some 75 centimeters per second. Phasing of a small compliance (volume) and large inertance (long pipe from engine to airbox volume) is different than of large compliance and small inertance even at the same resonant frequency design. In other words if you make a new airbox with different resonant frequency, you can still have at some RPM the very same phasing and performance as you did originally!

The stock intake is more like pure inertance design, so that at higher RPM the wave might not be even reflected back to the volume, so that no or less standing wave resonance could occur at high frequencies there, and it would be of less meaning since the intake lines are already long. This is not good, nor bad! It is a design choice for optimising distributed interrupted flow. It is good for making steady flow at the input, but it doesn't utilize compression sound waves very much. (to tell the truth, it does resonate but the low compliance volume causes such terrible Q of the resonance that one is led to believe the designers intended to make it quieter rather than a racing car.

Now if we have a look at the 13x6x2 inch box: the air inlet is not even streamlined, and that is the point. The area behind the air inlet tho the wall forms a volume that has little superimposed flow and gladly acts as a pressure buffer, also, the wave caused by the suction forms in a plane parallel to all other suction ports and perpendicular to air inlet.

Also, the extended suction hose (black plastic) is a complementing point: if you lowered frequency of the airbox, lowering the resonant frequency of the hose+air filter also makes sense. Plus, in my opinion, the stock air hose is rather short and could use a higher inertance than it has. (there is significant pulsing of air intake there at low RPM, try your hand or smoke or flame to see - it does blow air out! (not only sucks).

In my design, I would love to use gas diodes. LOTS of gas diodes, as I outlined in my airbox design thread.

The stock air intake just looks much like the intakes of old 1.2L 60HP petrol engines here. And guess, the OM617 without turbo is too a similar powered engine :).

As for any airflow design: it is really true that you make calculations, assumptions and design and make five, six, seven, eight, nine, ten variants to see which performs better. Sometimes those 5% additions count. The computation is NOT easy, as I have yet to find any useable CFD simulation software capable of simulating *COMPRESSIBLE* fluids -> gases. Fow water simulation is lots of software available. Recently (late 2007), Boeing or some other company used one of the most powerful supercomputers on the planet (BG/L platform don't remember ranking, from 1 to 3) to simulate just some segments of the burners in a jet engine. Not whole engine, not complete burners, just some of it. And it took some time too!

Where are the times when I made a perfect electronic design and assumed it would just work? :D

In fact the link it points to talks just about what I said above: a pure inertance inlet design. That is nice, but not the best. And that is just the design stock OM615, OM616, OM617 uses. The purely inertance designed intake is only resonant in the pipe connecting the engine suction to the airbox. Other sound power is wasted and - as I said before, running n-th harmonic resonance is inefficient too.

The speakerbox design all racing and street racing motorbikes use is the latest and most powerful air intake design. It does not rely on purely making more air available due to the pure inertance of the moving gas, but uses a comliance, a volume where to store the energy from the wave for a much, much longer time than it is possible in a reasonably long inlet pipe.

Actually the site proves one point: we need to increase the inertance or to provide gas diodes on the outside air suction pipe. A 2-meter long pipe to the air filter actually makes sense! WoW.

Where do get power to make the "compressed air" from?

You may want to ask: where does the power to make "compressed air" out of nothing come from? Good question. In an engine with non-overlapping exhaust and intake valves the power comes from the motion of the piston which cause the air to be sucked inside. More specifically it is the energy needed to make low air pressure in the piston volume which gets the mass of air into motion. For this reason the inlet pipe to the intake valve needs to temporarily provide resistance, and this is what the inductive coils do, but in our case of using air it is called an inertance. But as you can see the purely piston driven intake airbox resonator has the disadvantage of unfortunate timing and low rate of change. Sometimes you can see it corrected by opening the intake valve late, when the piston is well on its way down.

Overlapping valves - by that we mean that the exuast valve is still at least partially open while the intake valve opens. Understand that this happens in motion and at high speeds, so what it does cause is a draft. While the exhaust gases have been set in a very fast motion, also the exhaust pipe is long, it acts as a system with very high inertance that has already been set into motion. Where this system can be used: racing cars, motorbikes where emissions are not a problem, fuel injected engines, including diesels. :)
Note that this actually sets the air in front of the intake valve into motion faster and sooner than would the purely piston driven mechanism do, also the total energy invested into pumping/sucking the air can easily be greater, showing promise of higher air induction. And you can be sure that street bikes, etc use it at lot in their 2000-15000RPM range. Oh, did you note how this affects the exhaust sytem design? making the exhaust pipe too short, too wide or narrow would all decrease the inertance effect driving the air boost later. Now remember, we need the air to burn more fuel. Burning fuel into carbon and hydrogen is the least efficient thing we can do.


We will sort the exhaust manifolds in 4-cylinder engines into these categories:

1) All 4 exhaust ports connected very shortly into one pipe. This is typically used on non-overlapping exhaust and intake valves where the joined exhaust port feeds a turbo. Racing unfriendly and if used on normally aspirated engine a bad idea.

2) exhaust manifold with all 4 exhaust ports are connected with equal lenght pipes into a common exhaust pipe. This is the case of most stock exhaust manifolds. Nothing extraordinary. Just the bare stock. Car manufacturers use it also because the manifold can be made by sand casting of cast iron.

3) exhaust manifold "4-1" - The basic racing choice, you have four equal lenght pipes from the engine exhaust ports connected at the end into common exhaust pipe. The four pipes are long, rather straight with only the necessary turns, with diameter and lenght specifically selected for the engine, camshaft and RPM range.

4) exhaust manifold "4-2-1" - The best racing choice. Similar as above, but the ***finish here*** Properly choosing all six parameters (diameter, lenght) is not a trivial thing to design, some empiric experience and lots of calculation help. Web calc.

5) exhaust manifold "straight pipe for each exhaust port" - each exhaust port is equipped with a separate pipe, each of equal lenght and diameter. Usually straight. This is obviously a loud solution, and not always has only advantages over 4-2-1 configuration. But sometimes there is no other option, like with extreme engine displacements and/or extreme power levels (1000kW-8000kW). This simplest design can be even further improved by correctly using the entropy levels of the exhaust gases like in NASCAR exhausts (not discussed here).

Interrupted flow is bad idea

No matter how slow the fluid flow is or what the internal pipe diameter is, getting the fluid to move and stopping it is like moving and stopping a train.
1) you need energy to get it moving
2) when you get it moving at constant speed it is only the aerodynamic drag you spend power into
3A) when you try to abruptly stop it (by an obstacle) it either compresses the obstacle or the train or both
3B) while moving, it has inertia that can be used to either push or pull other trains.

Case 3B application: exhaust manifold. What 4-2-1 exhaust manifold does is that as one train leaves the station 1 it just catches the last bandwagon of the train that left before from station 2. It does not not crash into it (and thus slow down), they simply join into one train. Similar thing happens as these two trains join the extended train departing from stations 3 and 4. [[I cheated a bit in this example]]

The reason we do it

When you have a pulsed fluid flow; especially compressible fluid like air or exhaust gases; you have to remember that when you start pushing the fluid into a pipe everything goes fine until the flow finds an obstacle which includes diameter change and pipe end. As that happens a sound wave is reflected back. Did ...Imeant the pressure frontwave

I don't have an easy to understand example, the easiest example is the exhaust of a 2-cycle engine. If you don't know how it works use this approximation. Take a glass bottle. Put it under ... and .. water to fill it rapidly. as the bottle will get filled the water will suddenly strike high into the air, but after that the flow will be again smooth..

If you have read it this far, please leave a comment so I have a guestbook here. Thanks.

"Anodised" aluminium, alumina over aluminum pistons

Friction. Friction is bad for engine performance. Aluminium piston in cast iron liner is a relatively long lasting and low friction combination.Aluminium puston in straight aluminium cylinder is a bad idea. Dies fast.

What is usually done is coating, like iron coated alu piston in alu cylinder (Porsche), or better solution, nikasil, kanigen or similar coating in the cylinder and alu piston.

These solutions are good, but technologically complex and not easy to implement for a hobbyist racer.

Mechanical watches of the past used texts like "17 jewels", what that meant there were 17 bearings consisting of ruby or sapphire. And indeed, steel over ruby or sapphire forms a very good bearing. Sapphire (and ruby) are minerals which re Al2O3, aluminium oxide. Guess what! Pistons are mostly from aluminium!

What can be done is that you can make the sapphire-like coating over most aluminium pistons! Standar method is anodic process where the aluminium part is submerded into sulphuric acid (H2SO4) or oxalic acid (in my opinion far better for this purpose!!) and you turn on the voltage source for several hours. The result is a aluminium oxide coating that is somewhat porous (and this may help us) and should copy original alu surface. The problem is mostly in thickness achieved: it is small, micrometers, maybe up to 10, 20. Next problems is dimension change: the aluminium oxide CAN be cissolved in the acid too, so sometimes after the maximum thickness is achieved, new aluminium oxide is created only at the expense of dissolving some of the top layer.

In any case, the piston surface must be very smoothly polished to be of benefit to us. Sapphire is hard substance and any loose aluminium oxide would only serve as an abrasive. While still on the aluminium piston, it can be a very helpful ally, and while OFF the surface a very nasty enemy. So the mechanical surface preparation is of the most importance to us. Also, manu aluminium alloys can not be anodised! The last point is that the Mercedes pistons use steel inserts which don't like oxidising in acid at all, the steel surface would need to be well protected from the bath and fumes!

INVENTION: Caustic "anodising" aluminium oxidation process

I did that some years ago. Basic compound in the electrolyte is water and caustic soda. Yes, the very same caustic soda that dissolves aluminium and attacks aluminium oxide on its surface. However we can stop that process by applying electric current. I folgot the current density needed, but it was very high, optimal (for quality) temperature was 91°c-93°C, the bath also contained some dissolved aluminium in the process, clean fresh bath was a lower uuality producer. The other electrode was stainless steel. Lots of bubbling, bath temperature is kept by the current alone. One more point: I forgot which electrode was "+" and which "-" pole, it may not even be an anodising process then!

Disadvantages: The part diameter will be less than it was at the start of the process. Some aluminium alloys will totally spoil the bath.

Advantages: Many, it needs some observation, but it is very quick. The resultant coating is transparent, shiny with milky-white look. Very slick surface after the process depending on bath state can have high environmental resistance (TESTED! some of them were thrown out on ground, in contact with soil, rain and all the dirt you can get). After applying oil, the surface is oleophilic (attracts oil, yup, we want that!) and the friction factor on the surface drops a lot too. The material used were forged aluminium spoons and the surface of the later bath specimens had milky pearl-like look and after oiling kept incredibly slick surface due to the oil filled the pores.

INVENTION: Use elecroless ""anodising"" process

And now for something I do in the recent two days. I made a path and submerged a slightly worn aluminium piston. I cleaned and degreased it with acetone. Surface was polished with 6-micron SiC powder which also removed contaminants and aluminium oxide. After dipping the piston in the bath (some hours), a coating of grayish metallic color became evident in all exposed aluminium parts. The coating is hard and slick when tested with fingernail, not up to the slickness of the electrolytic coating above, but the thickness is also very thin. (it gets better as the coating gets thicker)

Maing ingredients besides water are two neutral organic compounds and a minority of some salts (weakly caustic). So far I could explain it only as that the compunds serve as an oxygen transfer agent, as I used tap water, and the sodium ions attack the aluminium to form hydroxide which gets mostly converted into aluminium oxide. (Aluminium hydroxide would cause the gray coloring for example). The two organic compounds would serve as an oxygen transfer helper, the aluminium itself acts as a catalyst for the reaction and the organic compounds propably help during the oxide formation to promote the formation of nanopores trough which the reaction is able to continue further down into the metal. To prove this hypothesis a presuurised container would have to be used and filled with air, say ordinary compressed at 10 bar (~11x the pressure we breathe) would be a good start. So far, I was adjusting the bath and the coating is thicker, slicker and darker. As a side note: with one liter of air you can turn 0.16g af aluminium into oxide (quick math, take care).

P.S. I am already improving the process, the result so far looks like a very slick colorless transparent sapphire!