Introduction and Part1
Based on my lifetime experience in metal fabrication and industrial installation around the world, the last 32 years before my retirement spent in working for a large manufacturer of heavy metal components in the departments of paper, mining and Hydraulic turbines, I am ready to discuss new proposals in aiming to improving the quality of manufacturing these components and even saving fabrication time in some of them.
The positions I have held in that last period were quality control inspector in the shop and abroad and technical director in the installation of grinding mills for the mining department.
I am not looking forward to a long time position but only to help interested parties to implement my proposals. These proposals are mostly related to customer concerns, to the solution of some of the problems, to introduce new measuring tools as well as determining the source of some others.
In order to avoid a long and tiring text, I divide it in shorter parts, learning the francis type turbine last. The last part involves most of my recommendations and is based on the user’s concern of having a mostly uniform and smoothly running machine,increasing the output and life of the part.
Part 1
: Measuring of a large taper
Around the end off the decade of 1970 or the beginning of 1980 I travelled to Arkansas to measure the tapered head of a paper mill since we were making a replacement. In our practice at that time we were strapping a sine bar piece in line with the axis of the part and if the measurement was the same on both ends, the taper part was to our standard. However, the old piece was made by another company whose standards were different you ours. However, the previous company machined two groves at a certain distance along the taper and measured it at the edges of the groves.
But the need to measure any type of taper without groves prompted me to device a method based on the fact that if a measurement is taken on two balls of the same diameter and known distance between them, and at 90 degrees to the opposite surface, the measured size is unique. By using simple trigonometry, the taper of the piece may be derived, This method became quite useful in measuring the taper on turbine shafts and runner hubs and saved considerable amount of time in achieving the required portion of surface contact between shaft and hub in a series of an order on tapered shafts and runner hubs. (US Patent 4777731A)
Part 2 Aligning motors to gear boxes
I worked in aligning motors to gear boxes on a large lamination plant in Brazil in 1962. At that time I followed the existing practice which today I consider wrong. Of course the platt worked but I suppose that it could have a longer and smoother life if it was done properly.
It is easy to explain by installing a dial indicator on a magnetic base, Then set the indicator to 0 while on the top of the pipe. No matter how tight the rods are on the magnetic base, by turning the pipe such as the dial indicator is at the bottom, the reading will not be the same. The further the dial indicator is from the magnetic block of the base, the larger will be the difference, this is something that even many qualified technicians afe not aware of. In the case of Brazil, it was even worse since the dial indicator was set on a flexible bracket around the coupling.
Part 3. Geometric tolerance
In order to match the runner hub coupling bores with those of the mounting shaft, a jig is used so that each specific bore on the one piece will correspond thru the jig to the same bore on the other piece.
There was a case where by placing the jig on the runners shaft case, three equidistant gaps showed up between the inner diameter of the runner hub and the outer diameter of the jig. This problem is related to the condition of the vertical boring machine and probably to setup, subjects that will be dealt with later, but it could have been previously detected, if the geometric tolerances were measured and in the suggested new way.
It is or was a practice to measure runouts on diameters by placing a dial indicator at the opposite end of the cutting tool. Since the distance between the cutting tool and the dial indicator is always the same, what is really measured is any ovality of the diameter, In this case the diameter was within 001’ in measurement and runout. The proposed new way is to use 4 dial indicators simultaneously, each at 90 degrees from the other. It is possible that the table of the vertical boring machine which is running on oil pressure is wobbling The same proposal is suggested for runout measurements on a flat surface
Part 4 setup and State of Machine.
I am taking as an example the case of the vertical boring machine where the runner of the previous casse (Part 3) was machined. A precision level placed radially on the table of the machine would have shown a discrepancy by running the table and recording the readings on a full turn.. Depending on the height of the machined surface with reference to the table, the results could have been predicted. It is also essential to verify that the rams of the machine run perpendicular to the table. We had a case that one of the rams on a VB machine was hit by the crane and the ram was running off course. I am not familiar with the process of repairing each machine but I can tell when something is wrong by seeing the result.
As the practice today is gradually replacing the cast Iron with steel weldments where the parts are more flexible, extra care should be taken in setting the parts on the table such as not to be distorted due to their weight or strapping. I have red about an old instruction on setting shafts that I haven’t seen followed.”set the shaft on tow V blocks at the two ends, set a dial indicator under the center until one of the ends is not touching the V block. Then with the indicator still in place, lower the center half the amount of the detector reading and support the center at that position. In the following part 5 I will come with samples on the field that suggest an improper setup during machining.
Part 5
Part 5 Improper setups for machining.
a)I was supervising the relocation of SAG and ball mills in British Columbia.That particular ball mill had a history of breaking the bolts on the connecting flanges at the center and according to our recommendations a company that was doing in place machining was called to machine the center flanges and more durable bolts were ordered in order to prevent them from braking again.
By connecting the cell to the gear, it was necessary to shim at two locations diametrically opposite by 1./16” more than the rest of the flange. To my opinion, the center flanges to be machined were not properly set and had sagged during machining or the ground was soft, or they were set on wood. The breaking of the bolts was due to the back sides of the mounting flanges had a recess machined but there was evidence rhat the square washers on which te nuts were tightening were partly on the machine recess and partly on the unmachined areas In a condition like this, the nuts would gradually get loose as if tey will be tightened on a soft surface,.
b) Always,After the complete assembly of the cell, hubs and trunnion on a new mill running the mill on oil, a runout on the trunnions was always present and close to the acceptable limit of 0.005”. This was because their centers of the trunnions were not exactly on the same axis. Most probably all mating surfaces not parallel to each other.
c) Never attempt to measure any heavy piece on the floor and be careful that even the cement may be moving by stepping on it without realizing, I have noticed it during theodoliting showing on the level change.
Part 6
Part 6. The effects in measuring in an uncontrolled heat environment.
On large metal pieces it is bound measurements to show slightly larger on the top than the bottom due to the air temperature being higher at the top compared to the bottom of the piece. However the effects are usually very small given the fact that heat travels internally on the metal.
Below are listed some cases where we were called at site to investigate and our findings:
Case#1: During installation, the bores on the head cover were found to be on a 0.020” radius larger than those of the bottom ring. The reason: It was winter and the installation was taking place in an environment that was heated and closed. (Isolated from the outside).
Case#2. The diameters of two identical pieces were found to be under minimum. In the shop, one was measured by me and the other by another inspector. I travelled to the site to re-measure them and I used the same tools as in the shop.The diameter on each piece was about 50” and each one was 0.003” smaller than reported. I investigated further upon my return. The table of that small VB where the pieces were machine was very hot when turning. The coolant at the cutting tool was keeping the temperature lower on that area, but the bulk of the piece was warmer.
Case#3 The shaft connecting the generator to the runner is usually made in two parts. In this case, one was made in Japan and the other in Montreal. We could have had other similar cases in the past. The parts did not fit on the spigots. Curiosity prompted me to investigate. I left a 17” inside micrometer into the hollowed part of the shaft for a few days. The diameter was measuring 0.0007” larger in the morning than in the afternoon. The reason, the micrometer was following the temperature of the piece at its contacting surface but the rest of the piece wa still warmer.
Part 7 Machining wicket gate stems on Horizontal lathe
I am in favor. of changes but only if they improve quality or save time. But some times, with the intension of saving time, overlook the consequences and the results affect the quality
It was our practice to add a counterweight at the rail of the gate to compensate for the weight difference of the nose of the gate. Someone decided to add the compensating eight on the plate of the horizontal lathe making thus things worse since the two forces were on opposite direction but not directly opposite to eah other and created further imbalance. Then besides the damage caused on the lathe itself , the diameters of the stem s were not uniformly round.
Part 8.New ways and tooling. Measuring spigots.
Spigots are recessions or protrusions on parts designed to keep assemblies on the same axis. In the hydro industry they are common on runner hubs. shafts, extension shafts and generators. Precise measurements are important so that all mating parts will turn on the same axis.
Because of their short height regarding the protrusion or recession, It is sometimes difficult to accurately measure them, especially since the micrometer spindle does not fit at all times in order to measure and more difficult when the horizontal lathe’s center point interferes. At times it has been necessary to.rely on transferred measurements at the expense of accuracy. I am suggesting a new method by using two blocks with extension pins that the measuring will be taken over the pinsThis device works well even with very large diameters for example with inner head cover spigots.
.
Part 9. New ways and tooling. Aligning of bores.
The practice of aligning bores, either measuring the alignment on the same piece (head cover for instance) or at assemble at sight (Head Cover to Bottom ring) is by placing a wire on the top bore, extending the wife until the other end. tight e weight at the other end and insert the weight on oil ar the bottom Then The wire is set center with the top bore and measure all other bores at 4 places, 90 degrees apart. This process is time consuming and degrees attention for the wire not to be damaged.
What do I propose: a) A large plate w;th a hole in the middle, the hole being larger than the top bore this plate to be set level using set screws and a master level.
b)A vertical assembly of tubes totaling the length required for the verification or the setup This assembly to bear on top another plate that will run on top of the first plate. For smooth running a thrust bearing. graphite or other means specified by engineering. openings to be provided on this top plate to see dial indicators set to run in the inner diameter of the top bore. Three bearings on an adjustable basis to be set to run on the inner diameter of the top bore.
c) Dial indicators to be se on top of the vertical assembly to ensure that the assembly runs perpendicular to the
original leveled plate. Other dial indicators to be set to read runouts in the other bores. The length of the assembly could be altered as required by adding or removing parts attached to each other by means of flanges.
Part 10. New ways and methods. Leveling bridge.
I am proud to present today what I regard as the most important element in my discussions. I call it “levelling bridge”, it is a simple structure and comes handy in setting up the bottom ring and head cover and with some adaptations in fabricating a Francis type runner.
It consists of a bridge type structure resting on three or four bearings resting on the face of the part and an equal number of bearings on the outside diameter of the part. On the top of the structure in approximately the middle is a flat plate with an adjustable inclination where a precision level is placed.
Let’s take the bottom ring as an example It is important that the ;piece is perfectly round: Due to flexibility when the piece is lifted up, care must be taken that it should be stored and lifted without distortion If distortion is detected it will show up during turning the levelling bridge around and it should be corrected by jacking as necessary.
Then turn the levelling bridge around a full turn, pick up the highest spot and equalize the opposite side to the same height. Do this at every 90 degrees and subsequently to the whole setup. At the end, run a dial indicator for verification. The piece is set,
This way the time is much less instead using levels and scope and the possibility of stepping on soft ground as previously discussed is almost non existent.
Part 11 New ways and methods.. The Francis type runner. Section A
This is a part where Hydro people always pay lots of attention.. With a good design and uniformly placed blades, it runs smoother, lasts longer and produces more power. Even if the increase of power is only 0.5% it is an amount of power to be considered., It also means longer life for other moving parts. Of course this is true for other type of runners as well, however I have spent a lot of time looking at almost every detail in the fabrication and machining of this type and in my opinion more attention heeded as compared with other types to produce a good part. Every little detail makes a difference.
Historically Francis runners were cast in one piece or two pieces, then machined preliminary the blades splined checked and corrected the inflow angles and radii, then applying a cutback on the outflow of the blades to provide an approximate opening within limits and finally welding up a recess on the blade with stainless steel and resplined finally before final machining. Later the blades were prepared on the floor within limits from templates and welded on a steel crown and band. Boxes were marked on the crown and band so that the 4 edges of the blade were set within the specified limits, two templates one radial and the other horizontal attached to a post were controlling the setting of the blade within specified limits and then followed the same procedure as previously explained. This way the position of the blades was controlled better and the blades were placed more uniformly than before.
Then completely. .machined blades out of stainless steel were introduced. Engineering developed a program that the shape of the blades could be machined within about 1mm to the design and accurately verified by the means of theodoliting. laser or other. What I consider wrong, is that at least up to the time of my retiring, the previous way of setting with templates and limits was used
In the next sections I will explain my ways in detail and show thar ir could be done effectively, faster, much more accurately and without any template.
Part 12 :New ways and methods: The francis runner fabrication Section B
Now we are ready to start the fabrication of the runner with machined blades, and premachined crown and band. Precision and details are important. Care should be taken on all aspects because the benefits are great if we succeed in keeping control of the process. Uniformity of blade setting will produce a better running machine, delivering more power, chance of a better balancing by adding less weight and saving time by reducing the material to be machined and lots of grinding and splining on the throat of the band. The amount of material required to be in order to allow in achieving the final dimensions will show up with practice
On previous Francis runner fabrications we were starting with the croup down on steel embedded in cement plates, the crown resting on machined blocks of equal height and Jevelled, while the band resting on pads attached to 6 columns, a post attached in the center of the crown vertically with a radial and a horizontal template set properly and the band adjusted to the center with set screws and the dial indicator Boxes were marked at the four corners of the blade (crown and band) according to the allowed limits and inserting and setting the blades was starting
In this process we drop the crown approximately in place and place the pads on the columns at the same level by using a wafer level or other means Since the band is flexible, it is important to store and lift in place the band so that will remain perfectly round, It is also very important that the inner part of the band be machined to such a shape that will heave a millimeter gap from the machine weld preparation of the blade to any surface on the inside of the band.
Then the band will be ser level by using the levelling bridge as described ain section 10 The bearings 0n the levelling bridge will show that the band is round, Do not attempt to set the band round by jacking the O,D. of the band at the Inflow, Use a long rod if necessary or a jack. The horizontal screws on the columns to the inflow O.D. are only to hold the band in place just by touching it Then use an attachment on the levelling bridge with two dial indicators to align the crow vertically and horizontally with the band.
I have stated before that it is proven that the blades can be accurately machined so are any holes drilled on the face of the blade. I have worked and kept in my mind all details in adapting a light bed on the levelling bridge which will accurately position from backing up set ribs and also matching of the face side of the blades and the holes. I also have a solution for leveling the band from under by using ser screws moving up and down without turning. Now, after dividing the outflow of the bang equally, into as many places as the number we can start inserting blades in the assembly. The benefits of the introduced changes against the up to now practice will be outlined at the last section.
Part 12 New ways and methods. The francis runner fabrication. Section C
As this is a new way to fabricate a francis runner aiming in minimizing distortion, improve the quality and saving time, it should be monitored closely, at least in the first fabricated pieces and evaluate the process or introduce changes to further improvement. As I am writing these lines, new ideas are coming to my mind.
1) I suggest that some targets be established around the area or on the floor where the runner is being fabricated, especially if only two theodolites are available, so that the coordinates of these targets will be part of the same coordinated system some small homies, probably 4-5 to be drilled on the peripheries of the band and the crown on horizontal level.These holes will be not deep and will disappear at the final machining. Some other holes will be on the blades done during machining. This way, by measuring the assembly as a whole piece will be possible to monitor any distortion and the relationship of the blade in the assembly as compared to the true position, by turning the coordinate system the proper number of degrees from blade #1
2) Tack weld the blades and braced the blades with pipe from the outflow on the face to the back of ththe previous or the next blade, depending on the rotation.
3) Brace vertically with pipes.
Bracing needs evaluation since it could involve more cracking on the fillet welds due to additional tension and restriction in movement of the material.
4) Block weld and full weld
At this point I do not recommend putting the runner on rolls, as to my opinion causes more distortion.
I think this is the reason that in Brazil I have seen welding the runner vertically, with the crown down
And for the same reason the runner should be put into the furnace with the crown down
.
Part 12 New ways and methods. Francis Runner fabrication Section D
After heat treat the runner will be checked for distortion and shrinkage by theodoliting and analyze these results. I haven’t mentioned it earlier but some noles should have been drilled during machining of the blades on areas near the inflow and the outflow that may be seen by the theodolites and see how close the blades of the heat treated runner can match the original coordinates. It will be done with the crown down on steel blocks. Do not put the runner on wood because wood shrugs slowly under the weight. It is preferable to be done, on a turning table. If the results are acceptable then the runner may be released for fillet grinding and NDE.
After this time, a final theodioliting analysis will be done, and the theodolite operator or engineering may decide that the runner may go to final machining or premachining. According to the results will be decided if the runner will be set with the crown flat on blocks or slowly tilted to suit better the design requirements The machining of the face of the band on the inflow will be marked by the theodolite operator., A few holes 4-6 could be drilled and tapped and filled up with say 3/16” or ¼”brass screws in a way that will be driven tight to the bottom and ground flush to the blade surface leaving a screw drive slot at the side and a center point on the center. These centers will be thedodolited to runner coordinates and kept in records for future field blade repairs at the field, then, slot and center pont will be filled up with epoxy, For the future repairs a robotic arm coordinate machine may use them as reference to detect and compare to the design and apply repairs as required..
At the end of this stage all remaining holes on the blades may be welded up and ground.
The benefits of this process remain to be seen at the end It is expected, uniformity, accuracy and time saving.
It may improve balancing by requiring less counter weight due to more uniformity. More counter weight means less welding under the shaft face of the runner, which translates to less deformation on the shaft face of the runner. Another benefit could be, if the throat of the band is more uniform, it will save lots of time in grinding and splinking the inner part of the band with templates to equalize the band inflow lip.
My francis runner fabrication sessions are ending up here. I want to thank all of you who had the patience to be with me for most of the time. If any of you has any questions I will answer them. I may have a last.future session on things I have seen and found interesting or even I have seen but I am not familiar with.
I want also to stress the fact due to age and health limitations I am not looking for a long term position but I will be excited to help implement my Ideas or what you have read so far and I will travel anywhere in the world for this.
Everything presented here can be found here: https://www.worldexperts.com/what-specifically-i-am-looking-for
Amigos Brasileiros talvez nós nos encontraremos mais uma vez. No Brasil eu fiquei um homem e meu coração bate igual ao seu.
One of the last two items that I want to mention but I don’t know enough about to present a solution is the internal micrometer to measure long distances.
We used to set a micrometer to measure long distances in our gauge room on a long rail. Such a micrometer consists of several length parts of steel, 1” of diameter as needed to produce the required length of the micrometer. As a result of it’s weight it deflects in practice.
To measure horizontally and up to a certain length, we set it by supporting it only at the two ends. Of course there id a sagging but in measuring the part the same way as is set, usually with the barrel reading on the top, the results are acceptable. For longer distances on the horizontal measurement,we used a support on the middle while setting the micrometer and three supports (ends and middle) of the same horizontal height while measuring. The micrometer had a lateral deflection which we were trying to compensate by moving it left and fight I am talking here about distances of 200” and more.
On a vertical measurement we couldn’t do much.
I have seen a Micrometer for this purpose in China, made up of welded cylindrical parts of light sheet metal, with larger diameters on the center and gradually smaller towards the ends. The deflection, if any, is not noticeable and they are very light in weight. I am not sure if they have any ribs internally but they seem to be the solution.
What specifically I am looking for. Part 13B Conclusion.
The last item I want to mention here is setting the dovetail ribs on a stator frame.
I was never personally involved in it but I have seen it taking a long time with a center post, plum bars, dial indicators and micrometers. Certainly from day to day and due to different temperatures or other reasons there are changes in the readings as compared to the previous days I want to suggest that two or three plates machined precisely and at the same time and set properly (I may make a difference. But this is after knowing more about it, I will be more confident to make suggestions for changes.
Here I am ending my presentation. I hope that some of you liked it. If you did and you are in a position to offer me a chance to implement what I am suggesting, I will be happy to do it and prepared to travel wherever necessary. I am not looking for a position to make up my living, nor for a 9 to 5 job. Just to help implement some of my suggestions I have already presented. Mainly because I believe in them and I don’t want them to be lost. I am ready to answer any questions you may have. I also think that my suggestions may be of interest to Hydro power producers white or before negotiating to place an order with a manufacturer.