BY-PASS or FULL FLOW?
Page 11
Page 12
Page 13
If you've noticed an error in this article please click here to report it so we can fix it.
That is the Que.stion,
By C. S. Le Clair, MIMech. E., M.Inst. P.
How Limitations of Space
and Increased Cost Determine the Degree of Efficiency of Engine Oil Filtration. Why Extravagant
Claims Made for Certain Types of Full-flow Filter Could Never be Realized
-FILTRATION is one of those everyday problems which appears simple at first sight but which, upon closer exaMination, becomes more and more complex. This is true of all filtration projects, from the .homely example of toffee brewing to the all-important prOblem of domestic water -supply. Due to the relatively high , and widely varying ' viscosities, the position is complicated still further when dealing with oil.. This article deals with 'the principles and methods of :ipproaeh, to the problem of , engine oil filtration, and an attempt has been made to -set these' out in such a manner that the operator will . be able to decide for himself which system is most likely to save him money and trouble by prolonging the life of his engine.
There are two objects in filtering the lubricating oil: (a) To prevent the working parts of the engine being damaged by abrasives particles, and (3) to prolong the useful life of the lubricating oil to the point where its replacement is necessitated by chemical decomposition and contamination.
Cause of Bore Wear The first effect of engine wear is usually that a rebore becomes necessary, and this frequently occurs much earlier than does the demand for bearing renewal. Cylinder bore wear, under conditions of correct lubrication, is due largely, if not entirely, to the presence in the lubricating oil of relatively large abrasive particles which are slowly pulverized by piston-ring action. The obvious solution is to prevent such coarse abrasive particles from finding their way on to the cylinder wall; it is true to say that one large grain of siliceous material would do infinite)) more harm than millions of particles of colloidal size.
In passing, it should be explained that when the abrasive particle-size is of the same order as the size of the pores in the cast-iron cylinder they are more likely to settle into the poles and not cut the metal, whereas particles in excess of -this size are bound to do harm. The well-known example of this is in the process of . lapping, where, if the lapping cornpound be Of very .fine texture, the cutting action practically ceases and results in a polished surface being produced.
Protective Oil Film
In a bearing which is heavily loaded and on which the speed of rotation is slow, the presence of oil on the metal surfaces has the effect of modifying the relationship of those surfaces by a physical-chemical action, and is covered by the term "boundary lubrication." At moderate loads and relatively high speeds, such as obtain in the bearings of a lorry engine, the bearing surfaces are separated by an oil film which is built up by the relative movement of the surfaces in the presence of a copious supply of oil.
The film pressure holding the surfaces apart has its origin in the relative movement of the bearing surfaces, and has little, if any connection with the pressure at which the oil is supplied to the bearings. Indeed, it is the designer's object to introduce the lubricating oil to the most lightly loaded part of the bearing to allow this phenomenon to develop naturally.
Attempts have been made to measure the thickness of the oil film which 'separates the bearing surfaces,' and it has been found extremely difficult to do so, but the modern. engine designer frequently allows-ark initial clearance of as much.as .002 in. in the bearings of a typical road vehicle engine. Thus the oil film may be as much as 0.001 in: thick. This condition will, of course, Vary With changes in load, speed and oil NiScosity.
Consideration will show that an abrasive particle of dimensions less than the thickness of the oil film between the bearings surfaces, can pass through this film without ever touching both surfaces simultaneously, and would, therefore be • incapable of doing any damage. A particle in excess of this size, however, will be dragged between the surfaces and cause wear Usually it does this by digging into, and being held by, the softer metal and subse• quently , cutting the harder metal. Ultimately, the particle is forced out of the softer metal as new and larger particles -enter the bearing, and it leaves-behind a minute crater.
What the Particles Do
Successive particles each have the effect of removing a minute quantity of metal from the harder surfaces, usually the crankshaft, and subjecting the softer bearing liner to a process of destructive pulverization. This effect can be readily seen on dismantling a bearing, when the crankshaft will be found to be polished, although worn, whilst the soft bearing-metal will present a matt surface which, under the microscope, presents a picture• of millions of minute craters, some of which are holding silicions or hard metal particles The perfect filter would remove all particles in the oil whether harmless or not, and the oil would remain clean and bright. The filter to do this can be designed; in fact, it has been proposed, but it would be nearly as big as the engine and would add several pounds sterling to the cost of the vehicle.
Perfection Can be Achieved
When the operator is prepared to pay the money and the vehicle designer can find the space, perfection in filtration can be achieved, but whether this would be worth while is doubtful as there are, in any case, other causes not concerned with filtration which limit the life of the machine.
We must, as in all design problems, come to a sensible compromise, and the solution lies in the direction of removing from all the oil which reaches the bearings and cylinder walls all those particles which are of such a size that they could cause wear. It would obviously be foolish to sacrifice the efficiency of this process in an attempt to remove from the oil all those infinitesimally small particles the only drawback to which is that they make the oil dark in colour, however reassuring it may be to the operator to see his engine oil clear and golden after many hours of use. Oil discoloration is almost entirely due to the soft carbon particles which are harmless.'
There are in use to-day two systems of oil filtration—the ``by-pass" or parallel-flow method, and the series c4 or " full-flow " method. The first takes a proportion, sometimes as little as 1/500th of the oil which leaves the oil pump and subjects it to a high degree of filtration, thereafter returning it to the circuit, while the remainder of the oil passes to the bearings with its content of abrasive matter undisturbed The second (full-flow) method, subjects the whole of the oil stream to a predetermined degree of filtration of such an order that no particles which are large enough to do harm to the cylinder or bearing surfaces are passed through the filter.
In practice, commercial examples of the by-pass, or parallel flow filters, are characterized by the fact that, for a time, they maintain the original colour and clarity of the engine oil, whilst having a very limited action in removing large abrasive particles which find their way into the oil circuit; usually only about 0.2 per cent. of such particles are caught before they reach the bearings.
Sources of Contamination
New engines contain a great amount of abrasive matter, however much care be taken in manufacture. It is released from the machined surfaces as grains of metal which become detached during the early life of the engine, and there may be traces of moulding sand used in the manufacture of the castings. Another source of contamination is the dust and grit introduced from dirty oil measures when lubricating oil is replenished, from grit on the hands of the mechanics, and the road dust which enters the engine through the breather pipe.
If the by-pass filter handles, say, 1/500th of the total pump output, then the oil circulates through the engine bearings possibly 500 times before a particular particle has been removed by the filter. To put it another way, a grain of sand has 499 chances of going through the engine before it is likely to be removed.
Dealing now with the full-flow or series-type filter, to keep this down to a reasonable size and cost, it is necessary to fit a by-pass valve. This allows some of the oil to bypass the filter element when the engine is started in cold weather, when the oil viscosity is hundreds of times higher than it is at working temperature. The effect of this provision is not serious because the engine, if cold enough for the oil to he by-passed, must have been standing for a long time, and abrasive particles in the oil will have settled to the bottom of the sump.
The oil on the bottom of the sump is the last to be heated,. and the foreign matter will still be trapped in a viscous bottom layer long after the by-pass is closed It should be remembered, also, that up to the time the engine was previously stopped, all the oil in the engine was fully filtered The by-pass valve is also necessary to protect the engine should the owner neglect to replace a dirty filter element.
Watch for Pressure Drop
There is one point which the designer must watch very carefully, and that is, if the pressure drop across the filter element under working conditions should rise to a figure approaching that at which the bypass valve will open, unfiltered oil will reach the bearings and the risk of damage is great. The correct filter arrangement, then, is one that is capable of passing the whole of the oil flow at a pressure which is only a fraction of that which will operate the by-pass valve and, in practice, a well-designed filter will have a pressure drop of the order of onefifth of that at which the by-pass valve operates and, indeed, only in the coldest weather does the by-pa53 valve open.
After many years of research and experience it has been found that a closely controlled wool felt is an ideal filtering medium. All sorts of other materials (metal, cellulose fibre, and so on) have been tried. but they all lack the fundamental filtering characteristics of felt. Filtering is by no means the same thing as sifting sand through a sieve.
It is not merely a question of having a material such as wire gauze with holes smaller than the particles one wishes to retain, because such a filter would become completely choked when a single layer of particles was deposited on the surface. Rather could it be likened to a barbed-wire fence which may have holes in it bigger than a man, but much more effective than wire netting having holes smaller than a man.
In the case of barbed wire, it is the entanglement of the wire which is effective and not the size of the holes. Now wool is the only readily available fibre which has a natural somewhat barbed surface, added to which it would appear to exhibit some eleetro-molecular phenomenon which causes particles, once, in contact with the wool fibre, to be retained even after being subject to severe disturbance.
Cleaning Felt Elements During the war, when economy in the use of wool was a national interest, Government departments and private 'research organizations did a great deal of work in attempting to wash felt filter elements to restore them to a clean condition. It was found, however, that even With powerful solvents such as trichlorethylene, . pumped at high velocity througl . the felt, particles could not be dislodged, and it was ultimately concluded that the felt.once loaded with particles, could not be completely cleaned by any process which did not result in its complete disintegration.
Another characteristic, of felt which is worthy of notice is that it presents an area much larger than the sheet of material from which the element is made.
It wobld seem that,, as far as development has gone, there is nothing better than a properly pro• portioned felt filter used in conjunction with an emergency by-pass valve, having an operating pressure much higher than the . normal pressure drop of the element.
Clear Oil not :Necessarily Safe The element size and rate of flow to be such as to retain all those particles which can damage the engine, whilst avoiding the misuse of the filtering capacity in the useless but attractive funct'on of producing clear oil: This golden stream may well be carrying large abrasive particles while the filter is loaded with harmless carbon Colloids.
Extravagant -claims have been made in the past for "full flow" filters, having elements of fine texture which, in _fact, are really " byTass " filters, as theelement, is quite incapable of passing the whole of the oil stream to the" bearings.. The accompanying 'graphs serve to illus trate bow.a filter of this type is really only a " by-pass " filter even when the oil is hot: