| Author |
 Topic  |
|
Doc
Keeper of the Scrolls
2010 Posts
|
|
 Posted -
28/05/2004
:
16:28
|
LANCASHIRE TEXTILE PROJECT
TAPE 78/AI/07 (Side one)
THIS TAPE HAS BEEN RECORDED ON APRIL 27TH 1979 AT 13 AVON DRIVE BARNOLDSWICK. THE INFORMANT IS STANLEY GRAHAM WHO WAS THE ENGINEER AT BANCROFT MILL AND WHO HAS BEEN THE INTERVIEWER ON MOST OF THE TAPES..
Pictures number 14 and 15. Neg numbers 777231 and 777230
I’m continuing now with the description of the two pictures numbers 14 and 15 in the Bancroft Folio. On picture 14, if you look on the low pressure bed just in front of the
cylinder, you will see a large tin, actually it's a two gallon stainless steel tin. This contains the reserve supply of cylinder oil for the engine. The cylinder oil came in 40 gallon barrels, and cylinder oil is of such a consistency that it requires to be warm before it can be used easily. What I used to do was have two of these tins always full of oil, one them on top of the high pressure cylinder where it warmed up slowly. And another one in reserve on the low pressure engine bed. This was mainly because it was just as easy to carry two tins as one when you were filling them up. It meant that
the chore of fetching oil from the top end of the engine house only had to be done about once every two or three days instead of once a day perhaps. There was also a kettle of oil set in a custom made depression in the insulation on top of the high pressure cylinder. This was very hot and so poured easily. The kettle was original equipment with the engine and fitted exactly into the aperture.
Looking at picture 15 you get a good view of the whole of the engine, down the inside of each cylinder. You can see the fish tank oilers on the flywheel shaft and the smaller, but similar oilers on each side of the second motion pulley at the end of the house. Notice the patch of white painted on the wall behind these. This was to make it easier to see the flow of oil into the bearing when you were going round checking the oils. If you look carefully at the left hand aquarium on the flywheel you can see, in the bottom window, the middle of the three oil oilers dropping oil into the bearing. You could see at a glance whether oil was running or not. This wasn’t a drip feed, there was a solid flow from each orifice into the bearing and an equal flow back from the pump.
You’ll notice that the top window of the aquarium is black. This indicates that the oil level is above the top of the overflow pipe. In other words the pump is delivering oil back faster than the oilers are taking it. This tells you that the pump is operating properly. I always kept the oil level high enough to support this condition as this was a fail safe device. Supposing you made a mistake and the oil feeds were turned off. You might have forgotten to turn them on at the start up, strangely enough, you might turn the oils off when you went to inspect them. The process of checking the oils became automatic and occasionally the brain switched off and you actually went into shutting down mode instead of oil checking mode. This has happened to me when I was distracted for some reason. The fail safe us that if this happens, as long as you have sufficient oil in the system, the overflow will lubricate the bearing even if the taps on the main oil feeds are turned off.
It’s very difficult to see on this picture but there was a pipe coming out of the fish tank oiler which dripped oil into small cups on top of the eccentric sheaves. This is quite unusual, these sheaves were usually greased, but on the Bancroft engine they were oil lubricated and the oil ran down into a trough below and thence back into the tank from which it was pumped back up into the fish tank. So these sheaves were being lubricated all day with a constant supply of fresh oil which was a good thing. I say a good thing but you had to be careful not to allow too much feed. It didn’t matter too much on the low pressure side but on the high pressure side the eccentric sheaves are next to the three cotton driving ropes for the governor. If there was too much oil dripping in these sheaves it could splash out on the ropes. They became soaked in oil and cotton ropes don’t like this. It encourages them to come to pieces and slacken. The governor ropes were soaked in oil when I took over and had to be replaced by Kenyons from Dukinfield who did any rope work we needed.
You can see that there are two ropes missing on the flywheel; reading from the left they are number 1 and number 5. From time to time ropes have been replaced on this flywheel and evidently someone felt that there was no need to replace those two. Quite properly because the engine wasn't delivering it’s full horse power for many years due to the fact that there wasn’t a full complement of looms in the mill. Bear in mind that each of these ropes is quite capable of delivering 50hp so 11 ropes coped comfortably with our load which was a maximum of 400hp.
In front of the flywheel can be seen the polished steel stand carrying three large gauges and the makers nameplate which you can read quite easily, William Roberts & Sons - Engineers and Millwrights - Phoenix Works – Nelson, Lancashire. This firm of course is now defunct, it went out in about 1948 I think when it was bought out by another company and the works closed down. In point of fact I think I'm right in saying that they made their last steam engine in about 1926 and these were exported to India. Actually they were some of the biggest engines that they ever made. Roberts weren't one of the big makers but they were a very good firm. The only fault with Roberts engines really was the fact that the steam valves only had single ports. This is a bit technical but I have decided that in these descriptions I shall give quite a lot of detail. If I go beyond your capacity to take it in just gloss over the technicalities. [As I digitise these tapes in 2003 I am conscious of the fact that I shall die soon and all this knowledge will be lost so please bear with me if I get a bit long winded. Remember that I will not tell you anything unless I believe it to be true.]
Right, single ported valves. This means that the valves had one large lot in them which coincided with a single slot machined in the port machined in the cylinder casting. This slot is the aperture through which steam is allowed to enter the engine. When you are setting the valves on a steam engine you have to decide how much ‘cover’ you are going to allow them to have when the valve is closed and at rest. It is this cover or overlap on the seat that actually closes the valve. It is the difference between the point where the edge of the valve shuts the steam off and the position it reaches when at rest.
If you think about the valve when it is closed it is held down on the seat by the steam pressure bearing down on it. If you think about a port size on Bancroft engine of say 16 square inches this means that at 140psi there is a force of 2,240lbs in old measurement, this is a ton bearing down on the valve. Imagine the friction between the face of the valve and the valve seat with a ton bearing down on it. This is not impossible to move, if properly lubricated the valve can be moved quite easily by use of the valve key when starting the engine. However, this cover has quite a significant effect on valve timing. Newton Pickles and I were in complete agreement that the engine ran best with the least cover possible. Most text books will recommend cover of up to ¼ of an inch on an engine of this size and some authorities even more. I used to aim for about a sixteenth of an inch. The way to set it was to open the valve with the valve key and listen for the point when the steam entered the cylinder. You can then alter the linkage until you have the position right.
So we now understand cover. The point about single port and double ported valves is that a double ported valve can be run with less cover than a single ported engine because by doubling the amount of surface available to the valve the events can be much shorter and retain accuracy. Burnley Ironworks engines had double ported valves and were always regarded as very good engines which ran well.
Picture number 16. Negative number 770300.
I’m still talking about picture 15 but as I describe the gauges mounted in front of the flywheel you might profit from looking at picture 16 which is a close-up of them.
The first thing to notice is that behind the gauges there is a backing plate that has one or two misplaced holes in it. This is not an original fitment, I fitted it to catch the small pieces of rope dressing that flew off as the flywheel was rotating and this plate stopped them being thrown back on to the engine. I’ll admit that I was probably more profligate in the use of rope dressing than my predecessors because I thought it did so much good and it may be that I was the first engineer to be troubled by flying grease.
This plate tells a story of its own. It was made of blued steel and the edges are protected by small bore brass pipe which has been split and fitted on to the edge. The original splash plates at the cranks are made in exactly the same way. It was a very common construction used by tinsmiths when making this type of guard or tray. This particular item was in fact originally used as a floor plate under a Barber Coleman knotting machine. Bancroft used to have two of these but one had been discarded and I found this plate in the cellar. It was exactly the size and shape I wanted so I used it behind the gauges.
Right, the gauges themselves. These three gauges gave the engineer essential information on how the engine was running. The picture was taken whilst the engine was running on a normal day with about 300hp load. I can tell this simply by looking at the readings.
The top gauge is showing the vacuum that the air pump is pulling measured in inches of mercury. It’s reading 23 ½ “ of mercury which isn’t a bad reading at all considering what a bad air pump we had. A good reading would be 25” or slightly more. The condition of the air pump isn’t the only factor that governs the amount of vacuum you can pull. Atmospheric pressure has a bearing on it as well. If you consider atmospheric pressure it can vary between 28 ½ “ of mercury and 30 ½ “, a variation of about 10%. This influenced the efficiency of the air pump. Another factor was the temperature of the condensing water which was being drawn in from the lodge. The warmer the water, the less efficient the pump. This used to be a big problem in summer where more than one mill was sited on a watercourse as the top mill was sending warm water down to the ones below. Therefore the best conditions were in winter when we had a bright clear frosty day with high atmospheric pressure. At times like these the vacuum was over 25”. Incidentally atmospheric pressure and air temperature influenced combustion in the firetubes of the boiler as well. Cold air at high pressure meant that the chimney could pull a better draught and draw a greater weight of air through the fire giving better combustion and a hotter fire. In the old days when we had coal fires at home we all knew this, in winter when the fire in the grate burned white hot we would say that it was ‘burning frost’.
Notice that the cock on the pipe to the vacuum gauge is in line with the pipe. This shows it is fully open and it always ran like this. The reason I mention this is that on the other two gauges I always ran with the cocks choked down to reduce the orifice to the gauges. This avoided vibration in the readings caused by fluctuation in steam pressure in the main steam pipe due to the intermittent opening and closing of the steam valves. If this wasn’t countered it could lead to accelerated wear of the spindles in the gauges. Choking the orifice in the pipe down with the clock put a damper on the movement of the needle and you got an average reading with a stationary needle. Notice that there is a complete circle like a pigs tail in the copper pipe to each gauge. This was not just a bit of fancy work, it has a serious purpose. As the gauge functions there is no flow through the pipe, simply the transmission of change of pressure or vacuum. However, water vapour gets into the pipe and condenses. It collects in the bottom of this loop and acts as a steam trap with air above and steam or moist air below. This means that water never actually comes into contact with the internal mechanism of the gauge and so there is no corrosion.
Notice that the face of the gauge is marked ‘William Roberts and Son, Nelson. And below this is a small trade mark. The trade mark is that of the Budenberg Gauge Company of Broadheath, Manchester and this indicates that they made the gauge for Roberts. The word ‘inch’ indicates the units to which the gauge is calibrated. All three of these gauges are constructed on the Bourdon principle. The essential element of the gauge which reacts to the changes in pressure in the pipe is a curved bronze tube connected to the pipe. As the pressure in this curved tube rises it tries to straighten out and this minute movement is transmitted via a system of levers to a geared quadrant which operates the spindle on which the needle is mounted. The only difference between the gauges is the thickness of the wall of the tube, the higher the pressure, the heavier the tube. In the case of the vacuum gauge the linkage was arranged so that the movement was reversed when compared with the pressure gauges as the vacuum or drop in pressure had to be represented in a clockwise movement. Obviously the pressure gauges worked in the opposite mode, it was the rise in pressure which moved the needle clockwise.
It’s worth mentioning here that there was another type of gauge mechanism, the Schaeffer diaphgram operated gauge. This was a very reliable and accurate mechanism but suffered from the basic disadvantage that the scale was not analogue, it was not composed of equal divisions. The diaphgram gauge was logarithmic and the markings started off very small and increased in size incrementally as the pressure rose.
[Originally Schaeffer and Budenberg came to Britain from Germany in 1914 and made both types of gauge. The bourdon tube ousted the diaphgram and the name of the firm changed when the government took over the firm during the war on the grounds that it was owned by aliens. When it was handed back after the war it was on condition that the name was changed from Schaeffer and Budenberg to Budenberg Gauge Company Ltd. I had this information direct from Richard Budenberg in 1987 when he reconditioned a lot of gauges for me at Ellenroad.
This is as good a place as any to inform you of a little known fact. The information came from Richard Budenberg. One of the major problems a gauge manufacturer has to address is to source a black pigment for lettering the faces of the gauges which will stay black over time and not fade. Budenbergs only ever found two methods that satisfied their exacting requirements. The most modern of these was ‘Letraset’ which was a dry process rub-down system of lettering first marketed in 1960. Budenbergs found that this was fade-proof. Originally they used nothing but a black resin obtained from China which when ground up with pure oil of Terebinth gave a black pigment that fulfilled their requirements. Each of the ladies who painted the gauge faces had a small piece of plate glass with some of this resin on. They would drop some oil of Terebinth on and grind it in with a small bronze roller and use this to paint the dials. Richard told me that the only way they could be sure of getting the best quality was for one of the directors to go to China and physically carry back a seven pound bucket of the resin. This was about a year’s supply and this happened right up to the firm closing in the mid 1990s. They still needed the resin because Budenberg’s practice was, if they were refurbishing a gauge they would renew the lettering in the style in use at the time of manufacture. So gauge faces were still being painted by hand even though many were done using Letraset. ]
I should note here that the gauges in picture15 show the engine at rest. This description is based on picture 16 where the engine was running.
The bottom right hand gauge shows the steam pressure in the steam main to the high pressure cylinder. It is reading a shade over 150psi. This is fairly high and indicates that when this picture was taken we were anticipating a big demand on the boiler. As this was late summer I don’t think it would be that the lights were going to be needed, more likely there was an expectation of a demand for process steam from the taping department. A normal comfortable pressure was 140psi. I have just thought of one further possible explanation. During the annual summer shut down, John Plummer and I used to prepare the boiler for internal and external inspection by the insurance company. The last part of this inspection was the ‘steaming test’. This was when the inspector returned during an ordinary working day and watched as the boiler was fired up to its maximum pressure under working conditions. The dead weight safety valve on the boiler lifted at 160psi and it sticks in my mind that this was what we were doing on the day when this picture was taken. If we knew the inspector was coming we would run at almost blowing off pressure until he came and open the stokers up to high feed when he arrived. It would take about 30 minutes while working to get the extra 10psi to achieve full pressure and lift the valve.
The bottom left hand gauge is the differential gauge. This indicates the conditions in the receiver between the high pressure exhaust and the low pressure steam valves. The difference between the reading on this gauge and the reading on the vacuum gauge is the effective pressure at work on the low pressure cylinder. It is reading 1 ½ psi on this picture. 23” of vacuum is approximately –12psi so the effective pressure on the low pressure cylinder at the time this picture was taken is a shade over 13psi. Recognise that this reading on the differential gauge is back pressure on the high pressure cylinder as well so the effective pressure on the HP is pressure in the main steam pipe, the right hand bottom gauge, minus the pressure in the receiver.
This indicates the balance of power being taken out of each cylinder. Remember that the low pressure cylinder piston is approximately two and a half times the surface area of the HP piston and so the power given by 13psi of steam in the low pressure is equivalent to approximately 40psi on the HP cylinder so under these conditions the LP is doing a quarter of the work of the engine. This was a normal division of labour when the engine was running on about half load. If we got to higher powers the differential pressure on the receiver could rise to approximately 30psi. At this level, the Low Pressure was working with an effective pressure of approximately 40psi. Back pressure had reduced the effective pressure on the HP side to 120psi given 150psi mains steam pressure. Under these conditions the LP was taking almost half the load.
One further thing about the pressure in the receiver. The engineer could adjust this pressure by making and adjustment to the hand wheel which we saw on the bracket over the Dobson block on picture number 14. This wheel adjusted the position of the linkage controlling the action of the wedges on the Dobson block and hence the cut-off of the steam valves on the LP side. If this cut-off was shortened it reduced the amount of steam allowed to enter the LP cylinder from the receiver and this raised the pressure in the receiver. This imposed back pressure on the HP and increased the pressure available to the LP thus increasing its share of the work done by both sides of the engine. In practice I never altered the linkage. The engine was quite comfortable running at this setting at light loads and automatically increased the share born by the LP as the volume of steam exhausted by the HP into the receiver rose as more load came on the engine.
That was fairly complicated but I hope you have managed to get the drift of it.
One more thing to notice on picture 16. If you look on the right hand side of the picture you will see the three ¾ “ cotton driving ropes which carry the drive from the pulley on the flywheel shaft to the governor.
Going back to pictures 14 and 15. These were evidently taken during the summer holidays. I say this because of the whiteness of the cotton waste and the fact that if you look carefully at the top of the relief valve on the Low Pressure steam pipe you will see that it is covered by a piece of cotton cloth. I used to drape the engine with fents during the holidays after we had cleaned up. This was to stop dust and dirt falling on the engine if a bird got into the engine house. A bird fluttering round the dusty girders in the roof could bring down an amazing amount of dirt on to the engine.
Let’s have a closer look at the evidence in these pictures and see what we can suck out of them. Look just the right of the Roberts name plate on the governor stand and you will see a small set of three steps. This was kept next to the governor and allowed the engineer to reach the linkage in the governor to oil it just before each start of the engine. The governors action relies on the precise but free fit of the pins in its linkage. This is best encouraged by precise manufacture and frequent application of small amounts of thin oil. The oil that you apply is thrown off almost immediately so a bare minimum was applied. We shall describe the governor in detail later.
On both these pictures you can clearly see the arrangement of rocking blocks which transfers the motion of the eccentric sheaves via the eccentric rods to the valves on each cylinder. Both sets of linkages are essentially the same as far as the cylinder, the only difference being the thickness of the rocking blocks to allow for the different positions of the valve gear due to the HP cylinder being smaller diameter. Once the rods reach the cylinders there is an essential difference, the HP side is controlled by the governor. The LP side has no connection with the governor.
The top eccentric rod on both cylinders controls the ingress of steam to the cylinder. In both cases it does this by driving the Dobson Block which is in turn controlling the steam valves by means of the wedges and catches incorporated in the block. On the LP side the position of these steam valve events is adjusted manually using the hand wheel above the block. On the HP side the adjustment is controlled by a linkage connected to and adjusted by the governor. The bottom eccentric rod operates the exhaust valves and in both cases the rod goes directly to the crank on the end of the back exhaust valve operating spindle. From the back spindle a separate, smaller rod links the back valve to the front exhaust valve. The effect is that as one valve opens the other shuts and the actual events are adjusted by the position of the circular valves on their seating. This is set by the position of the crank on the end of the valve spindle. The cranks are keyed on solid using a gib headed key and this position is set during manufacture. There are some small adjustments possible by means of threaded sections on the ends of the rods, these are for minor adjustments only.
Some words about gib headed keys and their fitting would not go amiss. The heads of these keys can be seen clearly just below the squared end of each valve spindle. The difference between a gib headed key and a plain key is the gib head. This type of key is used where it is not possible to get to the back of the key to drive it out when dismantling is needed to effect maintenance. The head on the key is shaped so that a wedge can be driven between the head and the face of the crank thus withdrawing the key and allowing the mechanism to be dismantled. Both plain keys and gib headed keys have parallel sides and a slightly tapered top side. They should be fitted dead accurate with both sides of the key bearing on the sides of the slots. The slight taper on the top of the key is to allow it to interfere in the fit and thus become solidly fixed in the slot. It is a common fault to find that keys have been badly fitted with no side contact and rely on the tapered section to give them a hold on the keyway. Such keys always come loose eventually. A properly fitted key with accurate side contact will not move. On larger sized keys and stakes the old fitters used to machine the middle third of the tapered surface out and fit the taper at the sides only. This cut down on the amount of filing needed to adjust the grip but did not damage the continuous side contact necessary for a good fit.
Looking at picture number 15, this is as good a time as any to talk about how steam gets into the engine. I’m going to assume that anyone reading this will know that the core principle of the steam engine is the admission of steam from the boiler under pressure into the cylinder where it exerts a force on the piston and thus converts steam pressure into linear motion. The con rod and crank convert this linear motion into rotative motion by turning the flywheel and this motion is transmitted by shafting to the machinery in the mill. We have been talking about the valves and how they are driven and timed by the motion transmitted from the eccentrics on the flywheel shaft by the eccentric rods. We now need to look more closely at how these forces are managed and controlled.
There are three phases to understanding exactly how the engine works. The first is the control of the steam delivery and how the engine uses it, second the action of the valve gear and third the control exercised by the governor. The last two will be described when we have more detailed pictures of the valve gear and governor. For the moment we’ll concentrate on the first, steam delivery and how it is used.
Steam is produced in the Lancashire boiler by burning coal and is delivered to the engine via a six inch internal diameter pipe from the junction valve on top of the boiler into the engine house. This steam main is installed so that there is a slight slope back to the top of the boiler. This ensures that when the column of steam in the pipe is at rest when the engine is stopped, any condensation in the pipe runs back into the boiler. It’s important to realise that the insulation on the steam pipe isn’t 100% efficient so there is always some condensation on the walls.
The slope back to the boiler deals with condensation in the pipe as far as the point where the bend in the pipe turns down to the stop valve. Any condensation in the vertical section above the valve is drained by the warmer pipe which is inserted in the lowest part of the section of the stop valve below the seat on the live side. This means that when the engine is warmed before starting any condensate in this section is blown into the engine where normal drainage will deal with it.
So, we have a supply of live steam from a dry pipe above the stop valve. This stop valve is immediately above the cylinder and is a very high quality steel mushroom valve made by Hopkinson of Huddersfield. Hopkinson’s were the standard valve manufacturers when this engine was built and over years of experience had developed a special nickel alloy that they called ‘Platnam’ which was very hard and resisted corrosion and erosion equally successfully. It was used for the mushroom head of the valve and for the seating. These valves would remain steam tight for many years. I never did anything to the stop valve at Bancroft beyond re-packing the gland on the valve spindle to maintain it.
When the stop valve is opened it does not allow steam into the cylinder, this is the function of the steam valves. All the stop valve does is charge the steam chest on top of the cylinder with steam at boiler pressure thus giving the steam valves a supply of steam. As I have explained before, the engine is started by charging the steam chest and then opening a steam valve manually with the valve key. Once the engine has accomplished its first stroke, the valve gear takes over and from then on the valve events are automatic.
It is important to realise that the governor has no control over the engine at the starting stage. It will start to rotate as soon as the flywheel moves but exerts no control over the valves until it has reached its set operating speed. Until this point is reached the steam valves operate at full travel admitting the maximum amount of steam.
We need to step back here and consider another circumstance which has to be allowed for when starting. Remember that when we start the engine we are actually moving an immense amount of weight in the shafting all of which is supported on plain bearings. During their period of rest during the night the weight of the shafting tends to squeeze the lubricant out of the bearings. This means that the first task of the engine is to overcome this ‘stiction’ in the shafting and simply get the shafting rolling so that the lubricant in the bearing can start moving and do its job properly. Recognise that the colder the conditions the worse this condition is. The worst thing an engineer can do is to allow the engine to start under full steam maximum torque. This can lead to all sorts of problems like damaging the gearing on the lineshaft by overloading it. In extreme cases the power on the piston will simply shear the high pressure crank pin off.
The way we manage this is to use the stop valve as a throttle. We don’t open it fully before starting, we just crack it so that we can start with full pressure in the steam chest but as the steam valves can pass steam faster than the throttled down stop valve can deliver it the steam pressure in the chest drops and in effect the engine starts gently on low pressure. Once the engine is rolling over nicely the stop valve is slowly cracked open just enough to keep up a gentle acceleration. Once the engine is up to speed the stop valve is opened fully and left like that all the time the engine is running.
When the steam engine was first invented steam valves on cylinders were very simple affairs. They ran on a fixed linkage and so their events were incapable of adjustment. The only way of managing the speed of the engine or the power it delivered was to manage the flow of steam into the engine by throttling the supply down using the stop valve. In many early engines this was done manually, the engineer stood their all day and used his judgement. With the introduction of the simple Watt governor this task was automated by connecting the governor to an equilibrium valve in the steam pipe whereby it controlled the steam supply to the cylinder. As the engine slowed down the governor dropped and opened the steam valve thus allowing more steam into the cylinder. As speed rose the governor moved in the opposite direction and cut the steam down. This is known as ‘governing on the throttle’. It was a big improvement on manual working and for many years was the standard method of controlling steam engines.
There were various improvements to valves and valve operating motions over the years as pressures and duties rose and engineers pursued greater efficiency. The valve gear on the Bancroft engine represents the peak of efficiency that was reached after 1900. You’ll notice that I said that the stop valve was left wide open while the engine was running. Bancroft doesn’t govern on the throttle but on the cut-off.
As engineers searched for efficiency they realised that giving a piston full steam for the whole of its effective travel down the cylinder was wasteful. There are very good thermodynamic reasons for injecting a fixed amount of steam at the beginning of the stroke, ‘cutting off’ the supply and allowing the rest of the power stroke to be accomplished by the expansion of the steam in the cylinder. This is known as ‘expansive working’. The development of valve gears such as the Corliss and motions like the Dobson block meant that the events could be controlled to achieve this ‘cut off’ and control it.
This how the Bancroft engine works. Steam at full boiler pressure is admitted to the cylinder at the beginning of the stroke and the governor controls the speed of the engine by constantly monitoring the speed and adjusting the point of ‘cut off’ to maintain the speed required irrespective of load or steam pressure. The only time this system can fail in its task is if the steam supply falls to a point where it is incapable of providing enough power for the load. Proper sizing of the boiler and good management by the firebeater and engineer ensured that this didn’t happen.
In effect, the Bancroft engine could run smoothly with high steam pressure and light load by cutting off at as little as 10% of the piston travel. It could run on lower cut off points than this but became a bit lumpy. This is the most stringent test of an engineer’s skill in valve setting and general engine management, to have an engine running smoothly on light load and high pressure. This became very important when we were weaving out the shed and the number of looms dropped off. In the end I was running 6 looms at 140psi with no problems at all. Newton Pickles said some very nice things about it.
Right! I’ve gone on at some length about that but I think that you should now have a clear understanding of expansive working. We’ll leave the valve motion and the governor for later and pick out some of the other information in these two pictures. 14 and 15.
On picture 15 you can see what looks like a drip feed lubricator on top of the HP cylinder just above the back steam bonnet. It’s a lubricator but not drip feed because it is for forcing oil into the cylinder. This is a lever operated Lunkenheimer lubricator. Lunkenheimer was an American firm which started in 1862 in Cincinnati, Ohio as the Cincinnati Brass Works. In 1889 it was renamed the Lunkenheimer Brass Works and then in 1894 took its present name, The Lunkenheimer Company. They are still in business as specialised valve manufacturers. In the early part of the 20th C. American lubricators had a large part of the market in UK because their products were so good. The lubricator on the Bancroft engine is an original Lunkenheimer but similar lubricators were made by English companies like Kirkhams of Bolton. This lubricator was used for injecting cylinder oil into the cylinder before the engine is started just to give an initial slug of lubricant. The Lunkenheimer is connected into the regular lubrication pipes and by manipulating the small cocks on the lines cylinder oil can be injected into both valves and the steam bonnets.
There is a tray on the floor in front of the cylinder lined with waste to catch any drips of oil and water. Notice also the groove lined with cotton waste which extends all along the base of the engine beds. This to catch any oil that runs off the beds. There is a particular reason for keeping oil away from the base of the beds. As I have said before, the only holding down nuts which are dead tight are the two at the front of the cylinders on each side. The bed must be allowed to breathe, to move slightly, so that it can accommo9date the expansion and contraction caused by changes of temperature. This means that the beds are always moving slightly when the engine is running. Any oil that gets underneath mixes with grit and dust and forms an abrasive paste which will eventually loosen the bed even more. For this reason we tried to keep oil from getting underneath.
There was a fault in the Roberts design which made this loosening of the beds even more serious. The pillow blocks which accommodate the flyshaft bearings are held on the bed by full length holding down bolts that extend right down into the foundations under the engine. As I have said before, these were never tightened dead tight. Any wear under the bed worsened this condition and you could never be certain with a Roberts engine that the pillow blocks were tight. It would have been far better if the blocks had been fastened to the bed by short studs. They could have been kept dead tight then with no injury to the bed.
This is as good a place as any to describe how to ensure that the engine beds were mounted tight and level to the stone cap on top of the brickwork foundation. The best way to achieve this is to install the stone caps on the bed but not attempt to level the tops of the stones. Once all the stones are in place the erectors cut a number of short steel bars, perhaps three inches long. A datum hole was drilled into the cap stones in the centre ensuring that the bottom of the hole was exactly at the level where the finished face of the bed needed to be. From there, a series of holes was drilled all over the bed, as each was drilled, one of the steel bars was inserted and checked by spirit level against its neighbours. Eventually the whole of the bad was marked out and at this point the masons came in and dressed the cap stones down until the datum holes were almost erased. Once this was done you were sure that the bed was dead level and at the correct height. The engine beds would then be installed using steel wedges for the final adjustment. Once the beds were in place and the holding down bolts in place, the gap between the bed and the floor was filled with grout and the last task was to nip the holding down bolts down and cut off any protruding ends of the wedges.
If beds were installed like this the layer of grout was very thin. Roberts’ didn’t seem to work like this. Bancroft had a very thick layer of grout. The levelling of the cap stones can’t have been very accurate. This eventually led to the HP bed coming loose under the pillow block but never enough to require action to remedy it.
Picture number 17. Negative number 772042.
Picture number 18. Negative number 776921.
It’s time we described the governor and its action. This is not the original governor fitted tom the engine when it was new. Roberts’ favoured the Whitehead governor which is easy to recognise because instead of a bob weight the centrifugal force generated by the governor balls is absorbed by a large coil spring mounted on top of the governor. If properly fitted and adjusted the Whitehead governor is perfectly satisfactory. The Ellenroad engine at Newhey had such a governor and I never had any problems with it. However, the original Bancroft governor was never a success. When the shed first started to weave the workers had to run out of the mill several times because the engine overspeeded, ‘ran boggart’ as we call it. This problem got worse after the Second World War when Bancroft ran a ‘moonlight’ or housewive’s shift. This entailed running the mill for four hours in the evening for weavers who couldn’t attend during the day. About 200 looms ran on this shift and the governor was set for 800 looms during the normal working day. The only way the Whitehead governor could have coped would have been to alter its adjustment every evening and in practice this was impractical and so the engineer had to govern the engine by altering the throttle valve. In 1948 Brown and Pickles removed the Whitehead and installed the present governor which cured the problem.
This governor that you are looking at is a Lumb governor made at Elland. These governors were widely recognised as being the best obtainable. I certainly had no problems with it and it kept the engine steady with a very small variation in speed.
Right, here we go! I’ll attempt a brief but comprehensive description of how it worked. The basic action of the governor is that as the speed rises the governor’s balls fly outwards and as the hinged linkage connecting the balls to the governor are deformed the thrust generated is absorbed by lifting the heavy bob weight in the centre. The characteristics of the governor are adjusted by balancing the weight of the balls against the bob weight so that at the designated governed speed of the engine the horizontal governor bars connected to the sleeve at the bottom of the bob weight are parallel with the floor.
These governor bars are captured in a fulcrum point mounted on a bracket which is part of the governor stand. As the governor bars are raised they force the left hand end of the bars down. The end of the bar is connected to a rod which transmits the motion via a series of bell cranks and rods round the back of the stand to the small arrangement of rods mounted in the sloping cast iron bracket on the front end of the casting which carries the turned governor column. All I want you to do at the moment is to recognise that the motion of the governor is being transmitted to these rods. They are connected to the linkage on the Dobson Block motion but we’ll describe this when we come on to a good picture of it.
Looking at the governor again, it is fairly clear I think how motion is developed in the governor and transmitted to the Dobson Block. However, we have a few complications that we have to look at. The reason why the Bancroft governor was so steady is that a fairly large movement on the governor only made a small alteration to the valve settings. This meant that the governor was very sensitive but steady, it ‘hunted’ very slowly and with minimum alteration to the valves.
We have to understand what this ‘hunting’ is. The governor is essentially a dynamic control system. By its very nature it cannot govern at rest, it can’t stop governing. The load changes by minute amounts constantly as looms are stopped and started in the mill and the governor is detecting these changes and is adjusting the valves to keep the engine speed constant. If load goes off the engine speeds up and the governor cuts steam admission back to slow the engine down, if load comes on it does the opposite. This constant variation is called hunting, the governor is hunting for the correct valve setting all the time.
If there is something wrong with the governor like bad design, stiff linkages, wear in the joints causing lost motion or bad maintenance and lubrication the governor doesn’t react quickly enough and by the time it applies a correction the condition has deteriorated and it has to make a big correction. This takes it beyond the point it was aiming at and it has to correct in the opposite direction. In real terms, the governor is swinging up and down wildly trying to achieve the correct speed but never gets there because it is reacting too slowly and too late. This is called hunting. Under these conditions the engine speed varies widely and the situation is unstable because if a major correction is called for the governor can’t cope and you either get a severe underspeed or the engine loses control and runs boggart. Many older type governors had a damper fitted on them, usually a piston working in an open ended cylinder full of oil. This smoothes the motion of the governor out but was an admission of defeat by the designer.
The Lumb governor at Bancroft had none of these faults. When running under normal conditions the governor bars appeared steady, the hunting movement was so small as to be undetectable to the eye.
We’ve already noted that one of the reasons why the governor was so accurate was because it was geared so that a considerable movement of the governor only made a small alteration to the valves. Putting it another way, the governor was only affective across a very narrow range of the total designed power output of the engine. When the engineer started the engine in the morning he set the governor by experience to the load he expected to have to provide for. The governor could cope with any variation in the load at this setting over a range of say 25hp. This adjustment was made by altering the length of the vertical rod which connected the left hand end of the governor bars to the linkage connected to the valves. During the day the load could quite easily vary by over 100hp so it looks as though the engineer is going to have to pay the penalty for having a sensitive governor, he is going to have to stand there all day altering the length of the governor rod.
Luckily for us, a man called Wilby recognised this need and devised an ingenious mechanism to deal with it. This is the compensator. Look closely in the bottom left hand corner of picture 18. You can just see the large nut and the opposing threads on the governor rod used to make the adjustment but your view of the nut is partially obscured by a small mechanism with a ratchet wheel in view at the front. There is a small handle on the front of it. If you turn this handle it rotates the nut on the governor rod by a system of gears. This is how you make your initial adjustment when starting.
The ratchet wheel you can see is keyed solid on to the shaft and rotates as the handle is turned. Conversely, if anything moves this ratchet wheel it rotates the shaft and adjusts the nut. There are actually two ratchet wheels mounted side by side with ratchet teeth machined in opposite directions. Just to the right of the ratchet wheels you will see what looks like a pair of callipers with the points almost touching the wheels. On the bar extending to the right from this ‘calliper’ you will see a link rod vanishing out of the picture. This rod is connected to the eccentric linkage to the valves and whilst the engine is running is constantly moving gently up and down oscillating the ‘callipers’. These callipers are actually the pawls which operate the ratchet wheels. One pawl acts to speed the engine up and the other acts to slow it down.
Now we come to the clever bit which is such a complicated motion that the only way to fully understand it is to watch it while it is working. When the engine is running at the correct speed and the governor is comfortably within its range the pawls are held away from the ratchet wheels. However, if the speed drops as load comes on the mechanism engages the appropriate pawl and winds the governor rod nut a fraction thus compensating for the extra load. If the speed rises it engages the other pawl and slows the engine down. It’s magic to watch this while the engine is running, the compensator will put on one tooth, have rest and then perhaps take one tooth off. This compensation mechanism operates all the time the engine is running.
There is more! About half way between the compensator and the end of the governor bar you will see a small wing nut. This adjusts the position of a pear shaped weight which rests on the compensator linkage and is in fact the mechanism which controls whether the compensator winds on or winds off. The alteration you can make with this wing nut is the most fundamental adjustment on the governor. Altering this nut alters the speed of the engine. It says much for the power of control exerted by the Wilby compensator that the smallest nut on the engine controls the largest force. Many engineers never touch this adjustment but I used it constantly.
I have already noted that by careful attention to valve settings, greasing the ropes, achieving constant steam pressure and attention to detail I was able to increase the speed of the engine without upsetting the weavers and thus raised production and wages in the mill. This small wing nut controlled all that. Proper adjustment of it had a tremendous effect on the profitability of the whole mill. I have also mentioned how small matters such as atmospheric pressure affected the running of the mill. Another key factor was temperature and humidity.
The major factor in variations caused by temperature and humidity was the effect they had on the efficiency of the friction drives in the mill, that is the ropes connecting the flywheel to the second motion pulley and the leather belts connecting the looms to the lineshafting. When humidity and temperature dropped the ropes and belts tightened and when they rose, the ropes and belts slackened. So in hot dry conditions the looms ran faster, in cold damp conditions the looms ran slower. There is an optimum speed for Lancashire looms which depends on loom condition, sorts of cloth and other factors. This is impossible to calculate but there is a way of allowing for it. We are talking here about a minute but crucially important adjustment.
I had a secret weapon called William Lambert, commonly known as ‘Billy Two Rivers’. He was a weaver on the pensioner’s side, under the lineshaft in the shed. He had been a tackler all his life but after injuring his neck carrying warps on his shoulder he went back to weaving. What that man didn’t know about Lancashire looms wasn’t worth knowing. When I had got the engine settled down after starting I used to take a stroll into the shed and stand under the lineshafting close to Billy’s looms. As far as the weavers were concerned it was just the engineer listening to his gearing but what I was actually doing was watching for a signal from Billy. He would make a small concealed gesture indicating more or less speed or leave it alone. The reason why we did this secretly was because weavers were a funny lot and if they had known what we were doing they would all have wanted a say.
I would go back in the engine house and make a tiny adjustment to the wing nut on the compensator and after about twenty minutes would go back in the shed again. Billy would give me the thumbs up. I often did the same after dinner as well. In this way I was adjusting the engine to the shed and it must have been an effective policy because productivity went up.
I can almost hear the collective sigh of relief, at last he’s going to stop banging on about the governor. Hard luck kids, there is even more!
You have heard me talk about engines ‘running boggart’ and how dangerous this is. A steam engine has enormous reserve power and under certain conditions can start to accelerate and will not stop until it has reached a speed sufficient to burst the flywheel. Look at the insurance company records and you will find this was an ever-present danger.
There are many reasons for this happening but they all have one thing in common, the governor has lost control of the engine. The designers of governors recognised this and we now have to look at the safety features incorporated in the design.
The threaded governor rod is the most important feature of these safety mechanisms. If this rod rises for any reason, steam admission is increased. A common cause of this was failure of the drive from the flywheel shaft to the governor. Under this circumstance the governor balls lose their power, the governor bars drop, the rod rises and the steam valves go to maximum admission. This is a recipe for disaster. The protection against this is the fact that the governor rod is made in two pieces connected by a hook. There is a small fixed point on the governor stand so arranged that when the engine is running, a small catch can be dropped which almost connects the fixed point to the lever which disengages the hook holding the governor rod together. This cannot be dropped in place until the engine has reached its normal running speed. Once in place, if the governor moves at all violently due to overspeed or sudden increased load or loss of drive to the mechanism, the catch contacts the lever on the hook, disconnects the two halves of the rod, the lower part of the governor rod falls and the Dobson Block motion stops operating the catches which open the steam valves. The effect of this is to instantly shut off the steam supply to the engine thus rendering it safe. Notice that the crucial factor in this safety mechanism is that the engineer has to remember to engage the safety catch once the engine is up to speed. If he forgets to do this, the engine is potentially dangerous. Almost all runaway engines are caused by failure to do this.
On a well set up engine with a steady governor the safety catch can be set to a very small clearance, That is, it is very sensitive to any unforeseen circumstance. The trick is to set it so it will accept the largest normal event, say setting one of the tape machines on, but react to anything greater than this. There is a penalty here which we have to consider. There was one common event which far exceeded the normal safety range of the mechanism. This was putting the shed lights on in dull weather. The engineer was responsible for deciding when the lights were put on and it was done by throwing a breaker on the main distribution board at the top end of the engine house on the HP side. This imposed an extra 100hp on the engine, far beyond what the safety gear could cope with. The way I dealt with that was to disengage the safety catch, wind the Wilby compensator open about five turns and then run round to the board and throw the breaker in. If the firebeater was handy I would get him to stand at the board and throw the breaker on my signal. Done properly the extra load could be imposed with hardly any variation in engine speed. Of course, I had to remember to re-engage the safety catch.
We are almost there now, just one more safety mechanism to note. You will notice that there is a shiny box mounted on the governor stand partially obscuring the governor rod and safety catch and connected by a flexible conduit to a pipe with a junction box. Quite obviously an electrical connection. This is the trip gear, an independent mechanism whereby a hammer is suspended above the lever that operates the hook that disconnects the two halves of the governor rod. This hammer is held up by an electro-magnet inside the box which is supplied with DC current through a circuit that runs right round the mill. At regular intervals along this circuit were what looked like fire alarm buttons but they were marked ‘EMERGENCY ENGINE STOP. BREAK GLASS’. If the glass was broken on one of these buttons it allowed the spring loaded switch inside to open, this broke the circuit, de-energised the electro-magnet, the hammer fell, the governor rod came apart and dropped and the steam was shut off to the engine. This system allowed anyone in the mill who was aware of a dangerous circumstance to stop the engine no matter where they were.
I have a confession to make. I used to test this system occasionally at stopping time. Theoretically, if the system was working correctly, I could stop the engine by cutting off the supply of power to the circuit. It always worked perfectly, I cut the power and the hammer dropped. There was only one problem, it didn’t knock the catch out! I reported this and we played about with it but never got it to work reliably. In the end I ignored it because I could hear the hammer drop and so would have known that someone was trying to stop the engine. Reprehensible I know but this was real life and we had a living to make.
SCG/13 September 2003
10,038 words.
Back to Stanley Graham's Page
|
|
|
|
Stanley Graham (Engineer & Researcher) Tape 10
