In an indirect way that is what we're talking about. Volume and "ring time", or sustain, are at opposite ends of the continuum. Volume is a function, in this respect, of the speed in which the assembly reacts to energy inputs from the strings. More reactive, more volume (at attack). But the more reactive it is, by definition, the faster energy is transferred from the string and also, then dissipated, i.e. the shorter the sustain.(Readers Digest version).
One of the challenges for soundboard builders is to figure out where on the continuum they want to be and how to get there. In addition there are certain structural requirements for the assembly to not come apart or collapse under a load for which it's not adequately designed.
When we're discussing safety margins, one of the considerations is that the greater the safety margin, i.e the more substantial in terms of load bearing properties, the less reactive the assembly will be which will result in lower volume and greater sustain. Finding the sweet spot where we build an assembly that adequately supports the designed load while remaining reactive enough to produce adequate volume but not so reactive as to dissipate energy too quickly and compromise sustain is the great challenge. There are obviously other considerations but that is a fundamental one and at the heart of the discussion.
Thanks for participating.
Original Message:
Sent: 06-20-2017 01:04
From: Paul Klaus
Subject: Correcting Steinway B SoundBoard Stresses
Terrence: Thanks for info. I am a nearly retired tuner afraid of asking foolish belly questions. Volume and duration are 2 of the 3 main elements of music so I initially thought the before/after board numbers to compare are decibels and ring time as increasing one or both would appear to be the objective. I am surprised/dismayed to find detailed numerical data on previous unknowns such as MOE and MOI, confusing graphs with more than 1 y-axis but no basic comparisons obtained with a decibel meter or stopwatch. Thanks for your help but I increasingly think this is not for me.
Original Message:
Sent: 06-19-2017 21:16
From: Terrence Farrell
Subject: Correcting Steinway B SoundBoard Stresses
Paul wrote: "...wonder if ribs can/should therefore feather down to zero."
One certainly can do that. In fact, I have built soundboards with ribs that taper to zero without intersecting the rim. Whether a rib does or does not intersect the rim is a product of desired soundboard flexibility at the edge and what the grain angle of the panel is in that edge area. When the panel grain angle is roughly parallel with the rim, a thin rib end extending into the rim can go a long way in minimizing the tendency for the panel to crack compared to not having any rib in that edge area.
------------------------------
Terry Farrell
President
Farrell Piano Service, Inc.
Brandon, Florida
813-684-3505
------------------------------
Original Message:
Sent: 06-19-2017 11:36
From: Paul Klaus
Subject: Correcting Steinway B SoundBoard Stresses
David and Chris: Thanks for help learning. I used to think the ribs need a tight fit against the case because the board acts as a buttressed arch. Now I don't know and wonder if ribs can/should therefore feather down to zero. I also wonder if the board is thinned from the bottom, top, or both and why?
Original Message:
Sent: 06-19-2017 11:12
From: Chris Chernobieff
Subject: Correcting Steinway B SoundBoard Stresses
That was by far your most powerful post David. Great question! I'm waiting in eager anticipation to see if Del or Jim will tackle that one.
I too, question the idea that the systems are really that different. If that were true, I believe the physical differences would be great as well. As it is, the average person looking at a soundboard couldn't tell you if it was a CC board or the other. The first question would be "what brand is it"? And then make that determination based on knowledge of the history of the company.
It seems to me that if one was a load bearing system and the other was a "stiffness/strain bearing system, their physical structure would be quite different.
For example (and this is speculation), if it was about stiffness and a glue joint to obtain that stiffness in a lateral plane. Wouldn't we see thick(er) panels and numerous small(er) ribs closely spaced? That certainly would create more gluing surface area. And the stiffness would be controlled by the panel thickness and to a small degree the size of the little ribs. But that's not what we see. Both look like load bearing designs to me. Instead of numerous little ribs with a thick panel , there are a few big ribs spaced far apart with a thin panel. As you see in other load bearing structures.
Paul's statement highlights the perpetuated fallacy.
Paul stated:
Chris: As I understand it a CC straight rib board has no load bearing beams so such calculations may/probably lack meaning.
First "straight rib": It's not a straight rib anymore! It's a laminated hygroscopically made arch before it's glued onto the rim if done correctly. This means its new natural tendency is to be arched not straight. I think that makes it stronger, not weaker. Granted, the opposite may be true if it was glued onto the rim straight and dry, and hoping the hygroscopic forces crowns it up after glue in. Then in that situation, would it want to be straight? I have actually seen that problem with some boards i have removed.
Flat as a pancake loaded
Flat as a pancake with strings removed
But once removed from the rim. Crowned up immediately. So how does the theory of a "straight rib wanting to be straight" fit in that scenario? Obviously, a sabotage of improper installation a 100 years in the making. If any board wanted to be straight, that one would.
Second "beams". I have come to dislike this term, because it confuses and falsely associates with architecture.
Although not 100% analogous, the study of leaf spring design stirs up many similarities. Thin on the outer extremities, and thicker in the middle. Arched. The more load there is the thicker it's made(more leafs added).
It's a spring that is obviously made to move carrying a load.
Soundboards are wooden springs meant to move carrying a load.
You wouldn't want an overloaded spring, nor would you want an overly stiff spring. Both have unpleasant effects in your car and piano.
Thankfully, there is a nice range of "personal choices" in the middle between overload and overbuilt as demonstrated by David and I in the choices of design. I'm still leaning towards the direction that David is overbuilding slightly due to the dependance on species M.O.E.
But that will be left for the future ongoing fun conversation.
------------------------------
ChrisChernobieff
Chernobieff Piano and Harpsichord Mfg.
Lenoir City TN
865-986-7720
chrisppff@gmail.com
www.facebook.com/ChernobieffPianoandHarpsichordMFG
------------------------------
Original Message:
Sent: 06-18-2017 23:15
From: David Love
Subject: Correcting Steinway B SoundBoard Stresses
They certainly don't lack meaning if by that you mean that any random rib scale will perform like any other. Rib scales clearly matter, though other factors *may* add to the complexity.
I do think that Jim articulated the complexities well. But I also agree with Chris that the complexities needn't paralyze us nor do they mean that spread sheets that take into account load, beam deflection, MOI, etc, are useless or misleading. They are indeed tools.
In my own designs many things do remain relatively constant: emc at glue up, rib radii, panel thickness, panel thinning, grain angle, etc. so that it is easier to discern how changes in the rib scales impact performance. If you are changing every element of the design for each execution then certainly you will have a more difficult time figuring out what's what.
It should be noted that many, if not most, current manufacturing protocols call far a hybrid design of modestly crowned ribs and some compression, emc of 5% is not atypical. Steinway NY may be one exception but I cant speak with authority about what their current system is. Most soundboard makers target specific rib scales that are either calculated on spreadsheets or derived empirically which they utilize repeatedly with predictable results. In my view a combination of calculation with empirical testing certainly makes the most sense. But it's not a crapshoot.
I have posted the following question several times and never recieved an adequate answer. RC&S boards are typically ribbed at 6% emc (at 70 degrees about 28% RH). If those pianos are kept at 45% RH (dampp-chaser equilibrium) then they will have some compression. A bit less than one ribbed at 5% (23% RH) but not that much less. The difference between a panel ribbed at 4% (18% RH--a typical compression crowned board) and one ribbed at 6% (a typical RC&S board) is only 10% RH. Steinway suggests that their pianos be kept in a range of 40-60% RH. At 60% RH an RC&S board will have just as much compression as a compression crowned board at 50% RH. So it seems to me that all current soundboard making processes have some degree of compression and differences are only a matter of degree.
My question is: if calculating rib scales is important for determining load support for an RC&S board ribbed at 6% emc, why is it not important for one ribbed at 5% or 4% and what real difference would that make in the actual ribscales that are calculated or used? In my view the differences in rib scale requirements have been exaggerated and quite possibly led to my experiences with overly stiff assemblies in those cases.
------------------------------
David Love RPT
www.davidlovepianos.com
davidlovepianos@comcast.net
415 407 8320
------------------------------
Original Message:
Sent: 06-18-2017 18:04
From: Paul Klaus
Subject: Correcting Steinway B SoundBoard Stresses
Chris: As I understand it a CC straight rib board has no load bearing beams so such calculations may/probably lack meaning.
Original Message:
Sent: 06-18-2017 11:24
From: Chris Chernobieff
Subject: Correcting Steinway B SoundBoard Stresses
Well articulated? Sure.
If you want to practice the " art of over-complication", then articulation is important.
Stating that the calculation of the distribution of the load(s) is a guess and assumption, i find incredibly fascinating!
It's simply not true. It can be precisely calculated. The principle law was first written in 1678 "Ut Tensio Sic Vis" by Hooke, and here we are in 2017 calling it an assumption.
I've shown how to do it with modern and easy to get equipment in past posts. (ad infinitum)
Another principle overlooked when building a structure is the minimal safety factor. If you design something to hold up a 1,000lbs and it ends up only holding 900lbs, then so what. That's good.
The whole purpose of a speadsheet is to insure the opposite ain't true. Which is the case with so many old boards that just get copied. Why copy a mistake? That's not fear, that's just asking a logical question.
The real assumption here is the one that says a speadsheet is isolated from the real life board. Yes, if you make it so. It's not isolated to me, but an integrated element.
Dimensions taken from soundboard
Load values placed on soundboard
Deflection values taken from soundboard
How in the world is that isolated?
It's a tool. And like all tools, the human has to learn how to use it.
The beginner fears the sharp chisel, the craftsman treasures it.
Yet, the competitor criticizes the choice of color of the handle.
And when asked to see his chisel............disappears.
Happy Fathers Day!
------------------------------
ChrisChernobieff
Chernobieff Piano and Harpsichord Mfg.
Lenoir City TN
865-986-7720
chrisppff@gmail.com
www.facebook.com/ChernobieffPianoandHarpsichordMFG
------------------------------
Original Message:
Sent: 06-17-2017 13:02
From: Jim Ialeggio
Subject: Correcting Steinway B SoundBoard Stresses
<… if the load were more, then my percentage would be pushed even higher than 78% and the original rib scale would perform more poorly even than shown.
This comment illustrates one of my frustrations with the use of spread sheets (mine included). Spread sheets have to rely on information that is assumed, not proven. I have my own sheets, and have spent many hours on them, as Chris and David have. The difference between Chris' and David's load assumptions illustrates one of the problematic issues in spread sheet belly calcs.
We can accurately calculate tension per string, calculate tension per unison, and calculate overall tension. Then we can calculate how much load those tensions will impose of the board, given a defined downbearing scheme. However, the actual fashion in which those loads are distributed on the spread sheet is a guess. This "guess" is an unproved assumption, and defines all the down stream calculations. Change the load and load distribution assumption and it changes all the down stream calculated results.
Why is this assumption a guess?
Among other things, the bridge is a beam which curves far away from a straight line draw between its end points. This means that beam cannot support weight in a linear fashion. Actually it can't even support its own weight. It will instead want to rotate around its end support points. It wants to rotate into the portion of the beam which is farthest away from that straight line drawn between 88 and 21. These rotational and leveraged forces will be greatest in the first capo section, where the out-of-axis condition is greatest. This means loads applied to the bridge will be distributed and smoothed, to some degree, by the bridge, but the distribution of loads will concentrate in the portion of the bridge that has the greatest rotational tendency...ie the 1st capo section. Not only that, as my own measurements continue to counter-intuitively show, the non-linear distribution of loads, actually seems to deflect up in the very high treble. From a pivot area under the top strut, and the end of the high treble bridge, despite being densely populated by significant loads...from my own measured observations…the bridge tends to deflect up against the loads, not down as if deflected down by the loads. (That is why the very high treble can often tend to show more downbearing than intended.)
So, as David's comment illustrates we are making a central assumption as one of the first steps of the design. The assumed load and the distribution of that load is different from designer to designer. This assumption, among others, is what take this calculated process out of one-to-one calculation territory. Instead, the calculations become devices which help draw statistical trends. In the spread sheet, the act of quantifying parameters such as deflection/length, rib section, etc, allows one to visualize and draw statistical trends, given the fact that there are many unknowns in the system….they do not present a 1+1 scenario.
One work around statistically here is for a sheet designer to define their assumptions in such a way that they remain consistent from belly to belly. Statistically, if you keep your load assumptions constant, it can have the effect of reducing the value of the assumption to "1". Multiplying anything by one, does not alter the thing you are multiplying.
Karl has expressed an interest in a belly spreadsheet. When folks look to purchase a belly spreadsheet, they are often under the impression that because there are beam, tension and load calculations, the spread sheet presents a one-to-one calculated result. This impression, as I have described above, is incorrect. The reason you will probably never see a rib design spreadsheet sold, is that the sheet is a statistical tool which is only as useful as the user's assumptions, load assumptions, observed correlations in actual rebuilds and pre-rebuild conditions, and correlations with other factors of the complex composite system which are not even included in the sheet. These other factors include number of ribs, board thickness, diaphraming patterns, scallop designs, length of scallops, thickness of thinned part of the scallop, emc @ ribbing, string tensions if deviating from "normal" tension, bridge height x thickness, end-of-bridge phenomena as related to location of ribs, etc, etc.
The complexity of the system forces us to make assumptions, and to exclude parameters which cannot be included on our sheets. The users eye and experience is paramount in the use of the sheets. This is one reason why, though I still consult my sheets, I have jumped ship, and am using an empirical method to test the as-built conditions of the entire as-built system, before its glued down for good. Rather than working from the spread sheet as I used to, I am currently using the sheets after the fact, to help draw trends. I described, previously in this thread, the pre-load procedure I am using as my way to analyze the system pre-gluedown.
Actually, for someone wanting to get their feet wet in this area, the simplicity of the empirical method outlined can serve as a pedagogical device for beginning belly rebuilders. It will get you in the generous "damn-good piano" ballpark without having to negotiate the statistical difficulties presented by the spread sheet. Statistical and empirical both work…but pedagogically, for me, the empirical method is easier in the "how do you do this stuff" department, and easier on the brain.
------------------------------
Jim Ialeggio
grandpianosolutions.com
Shirley, MA
978 425-9026
------------------------------
Original Message:
Sent: 06-16-2017 23:51
From: Chris Chernobieff
Subject: Correcting Steinway B SoundBoard Stresses
Thanks for sharing that info david. One more point I am not sure on is what your deflection value is based on ( I think we have discussed this before). If it's based on the species M.O.E. (1,560,000) then I think that will give you a false read. This is the reason I test an original board before removal, in order to get a more realistic value and accurate prediction of movement in the deflection formula. I'm seeing values between 5,600,000 to 19,000,000 for soundboards ( I call it S.M.O.E). This makes sense to me because its a laminated structure that is glued to a border. The species value is ok in simple beam structures.
I'll look through your spreadsheet and charts and get back to ya.
------------------------------
ChrisChernobieff
Chernobieff Piano and Harpsichord Mfg.
Lenoir City TN
865-986-7720
chrisppff@gmail.com
www.facebook.com/ChernobieffPianoandHarpsichordMFG
------------------------------
Original Message:
Sent: 06-16-2017 17:54
From: David Love
Subject: Correcting Steinway B SoundBoard Stresses
Clearly we have different approaches and methods for analyzing performance. Were I using the original widths (which I didn't but just for comparison) and species on the ribs (pine in this case) I would have done something more like this. By your method 78%. The load is based on a 3rd degree polynomial.
Rib # |
Calculated rib load |
Working length |
Width |
Target Height |
H/W |
|
|
|
|
|
|
1 |
60 |
730 |
23 |
21.3 |
92.6 |
2 |
61 |
905 |
24 |
22.8 |
95.0 |
3 |
63 |
1030 |
25 |
23.7 |
94.8 |
4 |
66 |
1115 |
26 |
24.3 |
93.5 |
5 |
68 |
1050 |
26 |
24.2 |
93.1 |
6 |
70 |
845 |
26 |
22.6 |
86.9 |
7 |
72 |
655 |
26 |
21.0 |
80.8 |
8 |
73 |
520 |
24.5 |
20.0 |
81.6 |
9 |
72 |
403 |
24 |
18.5 |
77.1 |
10 |
71 |
343 |
23 |
17.5 |
76.1 |
11 |
67 |
270 |
22.5 |
16.0 |
71.1 |
|
|
|
|
|
|
|
|
|
299 |
231.9 |
78% |
Here's how those ribs would perform under load (panel contribution not included in this). Based on 15M radius.
Under the same load the original rib scale performed more like this--not so good. The tonally weaker areas of the original somewhat corresponded to what you see below. I did not want to duplicate this rib scale and did not.
------------------------------
David Love RPT
www.davidlovepianos.com
davidlovepianos@comcast.net
415 407 8320
------------------------------
Original Message:
Sent: 06-15-2017 23:39
From: Chris Chernobieff
Subject: Correcting Steinway B SoundBoard Stresses
Hi Dave,
Well that S&S O scale you shared was actually pretty good! A 71% profile too. The lowest i have seen from Steinway was a 73%. I would pretty much leave the way it is if it were me. I was able to get the bridge location measurements off of a photo of an O Board I found online. Not exact, but close enough. Also, the rib scale is pretty darn the same as an L with slight differences in lengths.
Here's a screen shot. Thanks again.
------------------------------
ChrisChernobieff
Chernobieff Piano and Harpsichord Mfg.
Lenoir City TN
865-986-7720
chrisppff@gmail.com
www.facebook.com/ChernobieffPianoandHarpsichordMFG
------------------------------
Original Message:
Sent: 06-11-2017 18:45
From: Chris Chernobieff
Subject: Correcting Steinway B SoundBoard Stresses
Mr. laleggio,
Your method of procedure is interesting, but i'm sorry to say i believe that it's technically and strategically flawed.
The first flaw is the idea of using deflection (or as you call it "Bandwith of Stiffness") as a means of determining rib dimensions. Since ribs have a symbiotic relationship with one another, it will be all too easy to remove material from the wrong rib, and end up with ribs of uneven stresses and false readings.
Second flaw is changing only the heights of the ribs, and leaving the widths to a predetermined size. This strategy is very limited, and will most likely take the overall rib design back to the original at best, will leave the rib scale more uneven, and most likely result in a high rib profile. My guess would be an over 90% profile which is seen in most horrible designs. In other words compounding any original mistakes.
I suspect this is why you have repeatedly failed to submit for peer review any rib data frequently requested by me and other techs.
BTW, your false accusation that I'm using fear, Is based on you misquoting me.
You said:
"These discussions tend to lean towards the sales department, and in Chris's case, this is clearly a thread in search of sales. Setting up a scaffold of fear, by implying as he has done, that rib changes in thousandths of an inch can are heard is demoralizing, and is simply not justified by the reality I experience. Keep in mind, also, that there major brands whose structural profiles are considerably stiffer than the limited bandwidth presented by Chris, and Steinway….."
I never said you can hear a thousandth of an inch change. What I said was a thousandth of an inch change can cause a 35-60 lbf difference in bending stress. And uneven stresses cause problems.
I'm not using fear tactics or trying to demoralize people. That's simply just ridiculous.
Just show us the data, leave the personal attacks out of it, and maybe we can have a productive conversation.
------------------------------
ChrisChernobieff
Chernobieff Piano and Harpsichord Mfg.
Lenoir City TN
865-986-7720
chrisppff@gmail.com
www.facebook.com/ChernobieffPianoandHarpsichordMFG
------------------------------
Original Message:
Sent: 06-10-2017 21:19
From: Jim Ialeggio
Subject: Correcting Steinway B SoundBoard Stresses
Terry,
I spent some time working this pre-load procedure out, as my interest is, starting with a very stiff structure, incrementally back off that stiffness, to see what the tonal limits are. So, I've been working out a pre-gluedown, pre-load procedure, that allows me to target actual as-designed deflection/load in the complete composite system…ie panel, ribs, diaphraming, scallop design, thickness and length of the scallop and proximity to the rim, number of ribs, complete full rim clampdown (as if gluing in the board-without the darn plate in the way!), and (importantly I think) as-built compression of the panel. Though I triangulate on the emc of the board pre-ribbing, experimental error is such that knowing exactly where the emc was at ribbing is only accurate to +/- 1 or 2 %. So this sidesteps that experimental error and reads the total as-built deflection/load/length of the composite system, at whatever emc it was actually ribbed at.
It also lets me plane down the ribs, which I install too tall as a starting point, planing as necessary to hit my deflection/length/load, with load applied where the actual string load will in real life be applied.
It consists of a set of 20 calibrated, spring-loaded go-bars. The go-bars have a window and scale which can be set between 20-40lbs each including the weight of the go-bar. given internal friction of the go-bar device the calibration is within 15%-ish I think.
You can see in the above pic a dial indicator beam, and yellow tape on the case top edge. Though this beam deflects when spanning the case, measurements are taken at rib locations along the bridges, and the placement of the beam and indicator precisely triangulated so the beam is always in the same place when taking a reading at a given rib. Pre-and post-loading readings are taken.
readings are repeatable to .002.
At this point after adjusting the rib heights, which along with the deflection reading includes a tap test, I use these numbers to set bridge height. Best downbearing setting I've had, but that's a side benefit. I want to know what the limits of the system are, and belly by belly I'm working away at the stiffness parameter. Also working other parameters, because I don't think its all about impedance...but that's beyond the scope of this post.
------------------------------
Jim Ialeggio
grandpianosolutions.com
Shirley, MA
978 425-9026
------------------------------
Original Message:
Sent: 06-10-2017 06:50
From: Terrence Farrell
Subject: Correcting Steinway B SoundBoard Stresses
Jim I wrote: "Because knowing how to "tune the board", ie mass load if necessary, not by default, but if necessary, knowing how to read that new board, in order to direct tone regulating choices, and fully pre-loading and proving actual deflection-to-load parameter pre-gluedown, takes the statistical poker out of the mix."
I've pe-loaded a board pre-gluedown also. I'm just curious how you physically went about it. I used go-bars placed along the long bridge. Prior to placement, I measured the pressure of the go-bar with a scale. It was a bit cumbersome - just wondering if there is an easier method.
------------------------------
Terry Farrell
President
Farrell Piano Service, Inc.
Brandon, Florida
813-684-3505
------------------------------
Original Message:
Sent: 06-01-2017 10:22
From: Jim Ialeggio
Subject: Correcting Steinway B SoundBoard Stresses
< While there is some forgiveness in most assemblies, the targets, ultimately, are not as wide as you might think and if you venture too far afield you'll produce something unexpected and quite possibly unwanted.
David, this is a value judgement. It may reflect your aesthetic and it may reflect that of some of your clients. It does not reflect my aesthetic or that of any of my clients, with the possible exception or one.
Here is my experience...
1-Pitch clarity, textural and tonal differentiation, volume at tonal break-up, and voice-like sustain require that the board's structure pass a base threshold of stiffness in order to support these attributes. Once the a new board structure achieves this base threshold, the bandwidth of stiffness a board can possess and still be tone regulated to a fine degree is huge. Boards of varying stiffness will present slightly different tone palettes, but, given serious tone regulation chops, these different tone palettes will lend personality to a piano that is, right out of the gate, pretty damn good, highly musical, and in demand from excellent players.
2-Board design is at least 40% tone regulation(probably more)…ie, not board design at all. This part of the board "design" puzzle is never discussed in board talk, but it is so central to the success of a new board it needs to be explicitly discussed. Getting any new board to perform at or near its potential, requires high level tone regulation chops...and, it requires Time. Time meaning time to mess around with those regulating chops on that particular board, to learn what that board's unique requirements are. The chops include Choosing an appropriate set of hammers, hammers that match that particular board's impedance, Shaping those hammers in a way that supports register specific tone, on that particular board, Regulating and phasing the actual impact event with anal precision, and Regulating felt density to a very fine degree. A real biggy in this tone regulation process is that it may include some backtracking to correct tone regulation parameters, which we may have initially misjudged in the as-built board. This ability to backtrack and revise when necessary, may be one of the most important missed opportunities which prevent a new board with potential, from achieving that potential.
This tone regulating aspect is so central to the "success" of any new board, I actually shy away from providing "board only" services, since the success of the "board" is partially non-board related, and out of my control. Interestingly, the only new board of mine that supposedly "didn't work" was my first board. It was not until I tore it out, redesigned it again, that I realized the "board failure" was in fact a failure in my newbie tone regulating chops, not of the board design…lesson learned...eyes wide open and observing, and has positively effected all my subsequent board work.
Even on high profile instruments, given the essential pre-contract "what to expect" discussion any project should entail, I am so comfortable, at this point, in the generously huge functional bandwidth of boards, that I have been comfortable and highly successful utilizing experimental rib parameters on high profile re-manufactures...why? Because knowing how to "tune the board", ie mass load if necessary, not by default, but if necessary, knowing how to read that new board, in order to direct tone regulating choices, and fully pre-loading and proving actual deflection-to-load parameter pre-gluedown, takes the statistical poker out of the mix.
These discussions tend to lean towards the sales department, and in Chris's case, this is clearly a thread in search of sales. Setting up a scaffold of fear, by implying as he has done, that rib changes in thousandths of an inch can are heard is demoralizing, and is simply not justified by the reality I experience. Keep in mind, also, that there major brands whose structural profiles are considerably stiffer than the limited bandwidth presented by Chris, and Steinway, and to some degree agreed to by you with the way you voiced your caution. Perhaps I am reading into your caution, and that is not what you meant. If that is the case, please accept these comments simply as clarification.
Pre-contract frank, possibly sales-killing communication with clients is essential...and that wisdom is valuable. But the tenor of the statement, to me, implies a negative view towards any approach which exceeds the structural bandwidth that satisfies a certain manufacturer's aesthetic goals.
Said another way, my experience is… that its not rocket science…and, as a rocket scientist has reputedly been quoted… rocket science isn't rocket science either.
------------------------------
Jim Ialeggio
grandpianosolutions.com
Shirley, MA
978 425-9026
------------------------------
Original Message:
Sent: 05-31-2017 12:42
From: David Love
Subject: Correcting Steinway B SoundBoard Stresses
I would be surprised if Steinway had no design targets for the rib dimensions. Actual rib dimensions do vary but it's more likely one of execution rather than complete randomness. Rib scales within models do vary somewhat but there's a general pattern. Using something like "tallest rib in the middle and decreasing by 1/16" inch as you progress outward' (paraphrase) is probably not that useful since it will depend a lot on what your starting point is and there are other considerations. Using beam formulas to determine load bearing properties is not that new an idea.
Rib calculations are meant as a reference point, not as an absolute reflection of what will happen once the ribs are tapered, the panel is glued on and the assembly glued into the rim with some amount of compression (even RC&S boards assembled and glued in at 6% EMC have some compression under normal RH conditions). Obviously, those other factors contribute to the reactivity of the assembly.
Soundboard designers and manufacturers have slightly different methods and targets for calculating rib scales. My method is somewhat different than Chris's or Del's or Nick G's but they all use similar means--as opposed to just copying what is there. Methods for determining and measuring load distributions, what smooth load distributions look like, choices about downbearing settings, panel treatments including grain angles, thickness, thinning, panel drying targets, all vary to some degree and will impact the outcomes of a particular rib scale. While there is some forgiveness in most assemblies, the targets, ultimately, are not as wide as you might think and if you venture too far afield you'll produce something unexpected and quite possibly unwanted.
It's not clear to me whether or not a company like Steinway had general targets but didn't mind if the executions varied somewhat as a way to discover empirically whether something different might be preferable. Even under the most well thought out formulaic approaches, there is an aesthetic choice that must be made and a decision as to whether or not the most beautiful looking formula actually produces something that you like. And even then, tweaking the assemblies with slight downbearing modifications is pretty common even among the most current and state of the art manufacturers, at least the custom ones. I have taken apart assemblies where the rib scales did vary by more than one usually finds (typically erring on the side of too light) and most of those ended up as structural failures. That suggests that the ribs do play a role in the panel's ability to maintain compression in a CC system.
Whatever your belief in the infallibility of your own formula(s), there's no substitute for actually building it and seeing what it produces and seeing what happens to that assembly over time. Unfortunately most of us don't really get to see the later.
------------------------------
David Love RPT
www.davidlovepianos.com
davidlovepianos@comcast.net
415 407 8320
------------------------------
Original Message:
Sent: 05-27-2017 10:06
From: Peter Grey
Subject: Correcting Steinway B SoundBoard Stresses
I highly doubt that, when the Steinway B was designed, they "calculated" much of anything. I believe it was 'seat of the pants' good piano building from the 19th century. It happened to turn out pretty darn good, so they simply didn't change it.
That's just an educated guess, also based on the standard Steinway response to a technical question: "That's the way we've always done it, and we know how to build pianos..."
So it should not really surprise us to find some less than perfect stuff in there when we start ANALYZING things with the kind of equipment and knowledge available to us today (just a dream back then).
It's like an MRI or CT scan...they ALWAYS find something "wrong" in there. This is not a criticism, just an observation.
Pwg
------------------------------
Peter Grey
Stratham NH
603-686-2395
pianodoctor57@gmail.com
------------------------------
Original Message:
Sent: 05-26-2017 06:40
From: Terrence Farrell
Subject: Correcting Steinway B SoundBoard Stresses
Thank you so much Karl for painting such a funny picture. Next Whitney spinet I tune - or maybe even a Gulbransen spinet - I'll think of your post! :-)
------------------------------
Terry Farrell
President
Farrell Piano Service, Inc.
Brandon, Florida
813-684-3505
------------------------------
Original Message:
Sent: 05-25-2017 19:26
From: Karl Roeder
Subject: Correcting Steinway B SoundBoard Stresses
I'm a firm believer that it's better to remain silent and be thought a fool than to speak up and remove all doubt. However............. I've been wondering for some time now what purpose is served by the increasing height and width of the ribs in relation to their length. While Mr Chernobieff correctly points out the discrepancies in execution one finds when measuring old boards, I was taught that the general idea was to have the longest rib be tallest and widest decreasing by a 16th inch or so in both dimensions as one moved toward either end of the bridge. The 16th inch figure is just an arbitrary number for the purposes of this discussion. Since, as Mr Fandrich and others point out, in a C.C. board the compression of the panel is responsible for the maintenance of the crown and the glue joint between the rib and panel is responsible for maintaining the compression; why then aren't the ribs uniform in dimension? For that matter, why are they so few in number and so widely spaced? I would like to thank Mr. Fandrich for his excellent series of articles in the Journal touching on these matters and giving us all cool thoughts to turn over in our minds while tuning yet another Whitney spinet.
------------------------------
Karl Roeder
Pompano Beach FL
------------------------------
Original Message:
Sent: 05-22-2017 14:35
From: Chris Chernobieff
Subject: Correcting Steinway B SoundBoard Stresses
Chart Legend:
Dark area at the bottom is the load on each rib
Vertical Bars are the size of the ribs
The Line at the top is the stress levels of each rib
Board 1:
Analysis: Uneven stresses and incorrect dimensions for ribs. Both stress levels and rib dimensions not in sync with load distribution. Profile: 82% Stress: 2,099 Psi Load 975 lbs 10.8 sq in.
Board 2:
Analysis:
Not quite identical to the first board. Some ribs vary in stress up to 200 psi more than in the first board. The stress levels and rib sizes are again not in sync with the loads. Profile: 83% Stress: 2,044 Psi Load 975 Lbs 11.0 sq. in.
Why copy those obvious mistakes when they can be fixed and improved?
Board 3:
Analysis:
Now the load, stress levels and rib sizes are in sync. As they should be. Profile: 74% Stress: 2,208 psi Load: 975 10.6 sq.in
Amazing results. Live demonstration at my shop June 6.
------------------------------
ChrisChernobieff
Chernobieff Piano and Harpsichord Mfg.
Lenoir City TN
865-986-7720
chrisppff@gmail.com
www.facebook.com/ChernobieffPianoandHarpsichordMFG
------------------------------