Introduction to dealing with large components
I am Dimitri Favvas. an expert by my own definition in measuring and fabricating large components. I am basing this claim on my long time involvement as a quality control inspector and field representative mainly in the domain of hydraulic turbines.
I have had the opportunity to observe and study the making of the above mentioned components and the reliability of inspection reports, identify problems and propose solutions. Some of my proposals were adopted, I never had a chance to prove some others and more were formed in my mind after my retirement. This is why I am here. To find people interested on my ideas and implement them.
In particular I want to introduce new ways in fabricating with more precision and repairing as well in the field, Francis type runners without the use of templates, new ways in measuring geometric tolerances, spigot diameters and aligning of bores (i.g. bottom ring to head cover). I also want to put an emphasis on setup for machining taking into account the flexibility due to weight of large pieces, a practice that is not new but unfortunately not followed.
More details are provided below, although I am keeping key details off, in case a future user will want to keep them for its exclusive use.
I am making clear however that I am referring to the practices in use by my employer at the time of my retirement.
My firm position in the trade is that while we may use the same tools, the way to use them makes the difference in the first place although new tools should not be excluded. Just think of new ways instead of being conservative.
Interested to see my profile?
OFFERING BETTER SOLUTIONS
We are privileged in inheriting the knowledge of the past generations and having the chance to apply further improvements. Improvements are possible due to progress in technology or by applying new ways in doing things. I tend to believe that there is always a better way that either not everybody is aware of, or nobody has thought of it as of now.There is also a chance that some of us are doing things as learned from our predecessors which are originally wrong or missed an important detail.
With the above in mind and a life experience in the fabrication and measuring of large components, paying attention to encountered problems and examining their cause I am trying to present below possible solutions and/or improvements beginning with a francis type runner because this may include a vast majority if not all the problems and solutions of most of the large components I will be referring to.
I consider that one of the most important elements of a hydraulic turbine is the runner. The runner converts the water power into a mechanical by causing the runner to turn and energizes the generator which is attached to the runner For illustration purposes I include the following video, mainly to avoid explanations referring to the names of the runner parts and/or other turbine parts: ://www.youtube.com/watch?v=skQNf5_61Ps
The hydro power companies pay great attention to the design and the compliance to it since a properly fabricated runner runs smoother, produces more power and has a longer life expectancy.
During my working life I have seen runners fabricated of cast iron in one or two pieces. These runners were never perfect in performing and required a lot of manual work, mainly grinding in order to comply to some acceptable standards at their time. Later they were fabricated with steel crown and band but still cast iron blades which needed again more work to comply to some standards but since the blades were independent. they could have been placed more evenly onto the runner. This was an improvement but their placement depended on templates, lots of grinding and the spacing between the blades was not the same on all of them but was depending on established limits, besides the fact that sometimes templates used on individual blades on the floor did not match the templates used during their placement on the runner. On all the blades up to now, a stainless steel overlay was to be welded on areas where the blade was more likely to be worn.
Then, the machined stainless steel blades were introduced. These blades have the advantage of producing a uniform and to the design machined blade but since there is a transition on the way of fabrication, some of the old ways are still used for placing the blades on the runner, i.e. templates, limits on locating the blades and a number of checks used during fabrication of the runner with cast iron blades.
I have studied and developed new ways for placing the blades faster and more accurately that will eliminate a number of checks, will produce a more uniformed placing of the blades at a considerable time saving. Of course I am referring to a period when I was involved and I am not aware of any changes that may have been introduced during my retirement or are in use by other fabricators.
Being in almost daily contacts with hydraulic components during my work, I have identified the cause of many problems and I propose solutions here which include better ways in dealing with the fabrication, sometimes deviating completely from the ones in use at the time, which will produce a better product in terms of accuracy and time saving in producing the parts, changing ways of measuring dimensions and geometric tolerances and identifying sometimes the reasons of measured discrepancies such as set-ups, temperature, state of machinery, ways of measuring and flexibility of the metals..
I am not going to elaborate in detail on all the proposed changes for two reasons: One that some of them will require further engineering involvement and the other perhaps a third party may decide to proceed in securing the application or the proposed tool for their own purposes.
My interest at this moment is mainly a moral satisfaction rather than financial. I am introducing myself as someone who has had a long time contact with certain kind of products, studied conditions and problems and derived solutions or suggestions including new ways which if applied will improve the quality and the cost of producing them. I am ready and happy to share my knowledge with interested parties and I will achieve moral satisfaction to see them functioning.
FABRICATION OF A FRANCIS RUNNER
In general, I strongly believe that in making any product it is necessary to know what is expected of it, how to make it and examine the available means and the up to now knowledge in making it. I will also add, think of new ways to make it. As the latest is part of my personal character and my ambitions and it is the reason that I am here today. Any apparently small detail may make a lot of difference as I will elaborate further below.
This is a different method from the previously used (at least up to my retirement and by the company I worked that will produce uniformly placed blades accurately and without the use of templates. Since the parts are made out of steel it is bound to be subjected to some distortion and care should be taken to minimize, if not completely eliminate any distortion. This applies in all stages of production including fabrication, heat treat and machining. It is understandable that the larger and heavier the part, the greater is the distortion. The largest runner I have seen, has a diameter of 10 meters and weighs 450 metric tons. I intend to return to this subject later when the machining setup will be discussed.
The fabrication will start with the crown down and the band up. Since the band is the most flexible part of the assembly, care should be taken during storage and lifting that when the band is set up (usually in 6 columns) it will be perfectly level and perfectly round. The water passage surface on crown and band to be very close to the design in order to facilitate the positioning of the blades For the final leveling of the band use a technique that will produce no friction at all between the band and the leveling screw.
A leveling bridge type device is a new device which may be used in other applications such as the installation of bottom ring and head cover. It consists of an H type frame rolling at this stage on the top of this setup as well as the outside diameter of the band. An adjustable leveling plate on top of the leveling bridge ls to receive a master level on top of it. For leveling the top of the setup, install a dial indicator at the one end of the leveling bridge and with the level on top do a full turn to find the highest point. then turn 180 degrees, split the difference according to the level and work back and forth until the two opposite ends are on the same level.Then work on other points on the posts until all points are on the same level. By using an extension bar and dial indicators, set up the crown concentric and to the band and at the proper hight
As per my experience, the machined blades are produced around +/-1 millimeter or so from the design. At the time of machining the face side of the blade, drill about 6 holes on the face of the blade Then make a frame-bed at the face side as follows:
1) The frame will be attached to the leveling bridge on runner coordinates and matching up those of the blade.
2)The frame must be attached to the leveling bridge in a rigid way such as to avoid distortion and having as contact points to the inner surface of the blade 4-6 plates corresponding to the holes drilled on the blade’s inner surface as previously said and 4 removable plates at the 4 inner corners of the blade, (2 at the crown and two at the band). Use counterweights if necessary to compensate for distortion. The relationship of the removable plates must be established by tack welding 4-6 plates said 3mm thick on the inner surface where the original holes are drilled, bearing the same diameter holes and exactly corresponding to the holes already drilled on the blade. (use pins of the same diameter of the blade’s holes.) In a similar way, tack weld 4 plates, said 25mm thick following the blade surface at the 4 corners of the blade (crown and band) Then build up the inner surface plate to match the tack welded plates on the face of the blade Then build up plates matching exactly the plates of the 4 corners at said 25 mm from the tack welded plates on the blades surface.. The idea is that the 4 plates matching the hole will be permanent on the bed of the blade’s face but the corner plates will be removable. In this case the plates with the holes will be permanently attached to the inner surface bed of the assembly by wedding them on the bed and breaking the original tack welds while the 4 corner plates (25 mm thick) will be removable but fitting exactly 4 guide plates permanently welded on the inner bed.
3) Divide the band at equal spaces as the number of the blades.
4)With the inner bed attached to the leveling bridge, position the bridge on the radial line of tone of the blades, insert the 25 mm plates at the 4 corners, tack weld them on the crown and band position and then remove the levelling bridge from position and enforce the welding, taking care not to distort them from their original position. The idea is to drop the blade into position and adjust it by using set screws to its final position which is, matching the holes of the blade with the holes of the blade’s face with those of the said bed when attached on the leveling bridge. After all blades are set, it is recommended to brace the blades with each other and also to weld supports between band and crown in order to minimize distortion during welding.. Then block weld the blades starting with diametrical opposite positions before full weld.
MACHINING OF A FRANCIS RUNNER.
After heat treat, it is understandable some distortion to have occurred as related to the original setup. It is important to set the runner for machining at a position that horizontally and vertically will be the best average of all discrepancies from the original setup at fabrication such that the final product (machined runner) will maintain the relationship to the design as close as possible.
Important elements to consider for any machined part are: the setup of the part, the condition of the machine, the proper tooling, proper measuring equipment, the temperature and the qualification and common sense of the machinist.
1) The best setup is to set the part for machining as in its functioning condition but this is not always possible. However, it should absolutely be avoided to distort any part from its natural state by forcing it with the aim to move to a desirable location or to secure it on a fixed location. Never leave a machined part on a soft ground or wood for long time in order to avoid permanent distortion and never measure a machined part resting on wood or soft grown
2)Since a runner is machined on a vertical boring machine and the table of this machine runs on oil, it is important that this table runs at all times on a horizontal level. To verify for any discrepancies I will run the machine with a master level on its table to ensure that there is no wobbling during operation. At the same time verify that the machine’s rums are travelling vertically to the table.
Properly verified micrometers to be used and to ensure about the temperature of the measuring tools, the temperature of the measured part, and the coefficient of thermal expansion.
Discrepancies from original measurements related to temperature have been traced to the following facts:
a)During installation at the site,the radius of the wicked gate bores showed bigger on the wicket gate bores of the head cover than that of the bottom ring. This was due to the installation taking place during cold weather conditions in Canada and the place was heated. By the temperature rising on higher levels, the head cover had expanded more than the bottom ring as being higher.
b)The diameter of about 16” was found smaller when measured at the site in two parts by 0.003” at the site as compared to the inspection reports in the shop. The measurements in the shop were done at two different times by different people but both pieces were machined on the same vertical boring machine. By investigating the cause it was found that the table of the machine when turning was becoming very hot. The outside diameter of the part was at the ambient temperature due to the cooling compound being used but the rest of the part was affected by the heat of the table.
c)In a test case, an inside micrometer or 17”was left inside a hollowed shaft for 5 days. The doors of that building were left open at all times. In the morning the diameter of the hole was measured 0.0007” more than in the evening. The explanation: During the night that the temperature was lower the exposed areas of the inside and the outside diameters were cooling first. So was the micrometer resting in the shaft but the rest of the piece was warmer. The opposite was happening during the day. As a result, the size was showing bigger in the morning than at night.
3) A jig on a male spigot used to make the coupling holes on the shaft face of the runner, showed 3 gaps between the spigot of the runner and the jig’s inner diameter, The runout on the spigot was the same from either the cutting tool side or opposite to that and the diameter was within 0.001″ at any place. The up to now practice of measuring the runout at the opposite to the cutting end side is proven no effective in this case. the table was wobbling and the practice of measuring geometrical tolerances should be revised. Although running the table with a master level on it would have shown some difference, I suggest that it should be a practice to use 4 dial indicators 90 degrees apart on the diameters and the top surface.
4)New and simple ways to measure spigot diameters without transferring measurements that may produce accumulated errors and axis alignment simpler (without wires} are also possible.
5) There are more applications of the “leveling bridge” used to fabricate the runner that may be faster and render the use of a scope unnecessary.
It isn’t my intention to go in detail without further incentives but I am available to cooperate with interested parties
REPAIRING FRANCIS RUNNER BLADES AT SIGHT
After periodical inspections at site, detected cracks mainly on the fillets are repaired and blades are distorted due to welding os other conditions. They are restored close to their original position by using templates and the opening check at the outflow of the blades. It becomes more difficult to use templates as the runner size increases and templates are more difficult to use. In this case I will advice to use some prerecorded spots on the inflow and near crown area on material different from that of the blades and then use a robotic arm to check coordinates of the blades and compare them to their original condition.
Finally I want to add a note that in the process of saving time during fabrication some of the operations may be done wrong as compared to previous practices that were done properly. An example is that I have seen during the machining of wicked gates on the lathe to place a counter weight usually placed at the thin part of the gate, placed on the lathe’s plate instead.