Friday, July 18, 2008

Hard anodizing in warm baths!

So, I developped hard anodizing process that does not require supercooled sulphuric acid. One of the keys are bath additives. The other is current density (with large surface parts, cooling may be needed, the third and most important is to make the process start at the surface. And that one seems to be the hard part.

It is easy to form a hard anodized layer ( a dark brownish-gray thick layer ) once it started forming and this may be the property that one german company uses to form 15 micron hard anodized layer in just one minute and do it selectively! Gramm Technik link here.

Theory behind hard anodizing (type III anodise) - You need higher current density and you need high voltage. Sulphuric acid has high conductivity at high temperatures which will limit your ability to reach any higher voltage in the bath. Also high current baths get warm very quick. So cooling has to be used, ideally to the point that the bath is ready to freeze. That makes 1) the acid attack the formed oxide less and 2) the intrictic conductivity of the bath is lower, it is easiet to reach high voltage.

I simplified the above a bit, but here is what I have done: I decreased the bath conductivity without diluting the 8% sulphuric acid at room temperatures. The ability of the bath to attack the oxide is also lower. Bad points: some reaction may allow appearance of self-catalyzed burning holes on the anode, these may eat the aluminium out without stopping. This can be cured by simply tuning off the current for a second, or by rising the part above the bath surface. Then anodizinfg resumes normally.

What I found out is that the surface of the aluminium is very quickly covered by whitish protective oxide layer after turning the current on. Initial current is high, but quickly falls off and the layer does not seem to grow substantially if at all. That is at 30V voltage! No oxygen bubbles on the anode.

I have to note hat I use only mechanically cleaned and freshly polished alu samples, no chemical cleaning, or HF acid cleaning (or nitric acid). After some time the first oxide will be eaten trough and hard anodised layer will start to form. This happens mostly at the bath-air interface and follows straight down - propapbly some kind of salt created there activates the surface. After that the hard layer spreads to the sides and some aluminium thickness is removed in the process. Anyway, the created layer is not scratchable by a nickel coated metal probe of the multimeter.

Next point of achievement was bringing the cathode closer to the anodized sample and the current went up fivefold to tenfold. I did that on anothers alu sample and the overall thickness went up from 0.3mm to 0.35mm! That means the oxide layer is far thicker than 50 microns! Again, little or NONE oxygen bubbles on the anode.

Tuesday, July 01, 2008

Air boosting driven by exhaust powered vacuum.

The basic idea goes like this: to make 20% boost on an OM617 engine at full RPM you need about 2.5kW of power in case of adiabatic compression. I used adiabatic process calculator and here are the numbers. 135 L of air compressed to 112.5 Litres from starting pressure of 98.94 kPa and end pressure of 127.71 kPa, start temperature of 24.34°C and end temperature of 46.85°C requires 2526 J of energy. Those are the parameters of air that contains 20% more air molecules than the one second rotation of 3.0 L OM617 engine at 4500 rpm. You see the air has far higher pressure difference than 20% (29%), and it is the result of increased temperature of the compressed gas.

The below assumes air density of 1.3kg/m^3

Now to the crazy point. Previous article explained that properly phased suction vacuum is the power behind the resonant oscillations. Maybe you have seen or heard of devices that create vacuum by using compressed air and ejection nozzle. If we make a similar device utilising the exhaust gases we may too be able to make a reasonably strong vacuum source. How should this help us boost air pressure to the engine? We can use it to suck air out of a long tube to make it move at, say 450 km/h speed (125 m/s) and divert part of this moving air by aerodynamic flap or wing to a longer intake tube where the fast flow will slow down to the ordinary suction speed at the expense of being compressed.

The complexity of the design and calculations is high, and benefits are only small, but it is a device with no moving parts and no maintenance. Plus, when properly designed it can be added to existing supercharger designs and to some extent also with certain turbochargers - those which still provide enough power on the turbine output. There won't be many such systems, and those use three turbocompressors in series already ;-)

What am I aiming to achieve then? 20 to 40 kPa boost pressure for the air filer inlet, which would amount roughly to 15-30% more air molecules. Remember that efficiency of this system is much lower than that of the turbocharger, but it has lower delays and you only need to change muffler bearings once in Platonian year.

The last advantage of this exotic boost system is this: past the divisor the flow that leads to the ejector can be equipped with a small tube to let in it diesel fuel to make flames out of the exhaust, or water (or water + something) to make experiments of other kind (like plumes of stink? or even improving emissions!) or glycerine mix to make white or colored smoke trails behind the bike. For many people, that would outweight the former advantages.

How do Ejectors Work?

Very informative air-combustion gas-water vapour gas ejector desigh considerations from 1956.

First thing that will help us understand how much boost we can getis from the Bernoulli equation. That one only takes STEADY, not pulsed flow into consideration - plus, flow of an non-compressible fluid, like water. But still it provides the basic measure - scale. If you enter the values you will be able to see how much is the kinetic energy compared to the energy already stored in the fluid, because air, as we breathe it is already compressed! Comprerssed to about 1 atmosphjere of pressure. The Bernoulli calculator is here.

Post Scriptum: My math tells me this idea is not that practical because due to practical limitations on the size and flow capacity of the exhaust I can forget about a 20% boost. One meter of 6cm diameter pipe flowing 350 L of fresh air per second carries the kinetic energy of 28 Joules. Very modest compared to the 2.5kW requirement I started with at the top. This can only be useful as a sealant of the resonant pressure peak.

Post Post Scriptum: I was mistaken with the statement above, the kinetic energy of the WHOLE mass of 350 L per second is 3486 Joules and that amounts to 3486 Watts. You just need to take care of losing much less in the pipe in aerodynamic friction. Next you will make the pipe long enough to carry just enough energy to hold any counterpressure pulses from the resonator.