If you're a serious tech and especially a piano rebuilder or restorer, you've come across the subject of string downbearing.
Where I most often wage battle with down-bearing of late is in the actual setting of string bearing during piano rebuilding. And I've arrived at a place of shocking heresy. And here it is.
The issue I've always had with building new soundboards, with the bridge work as well - is whether to complete the bridge work inside the piano with the board and bridges glued in - or on the bench. And here is my preference: to first set the bearing obviously with depth calibrations made inside the piano. But then, to plane the caps out of the piano as well as to mark, drill and notch the bridges at the bench. And then, to also pin the bass bridge. Then, to glue the bridges to the board. Then, to drive the treble pins into the bridge. And finally to glue the entire soundboard and bridge assembly into the rim.
The issue however is that with a dry-fitted soundboard during the down-bearing setting stage, a great element of risk and variation is introduced because a dry-fitted soundboard can present differently at total sagitta elevation differently than the final glued in board. Plus, one must take into account the EMC of the board.
So here is my heretical approach that I'm actually prepared to execute with some degree of confidence:
First, I'll pull out my 5% EMC target board and dry fit it, plus the bridges, pinblock and plate all into the piano. My board in this case has consistently pre-crowned ribs at a 60' radius, also bellied at a 60' radius press. I'll then clamp the soundboard to restrict it from leaf-spring flexing under compression where it could slide across the inner rim.
Then, I'll use the same I-beam that my plate hoists run on, to support overhead wood blocks running down to machinist jacks that contact the new overly-thick bridge caps. I'll then create a very modest amount of compression on the long bridge about where it crosses the widest rib and span of the board. What this will effectively represent is the future downward weight of the strings under tension, generating downbearing. While I will expect 450 lbs of weight from the strings, I will not generate quite anything like that, but instead be conservative, only pressing the board through a relatively flexible phase until it reaches a degree of stiff resistance.
Next, I'll draw my string across the bridge in the traditional way of setting downbearing - and the only gauge I will use will be a simple dime. Yes, it's heretical. I will then disassemble and plane the bridge surfaces down to the saw grooves I made with string and saw. Then, I'll put everything back into the piano and check again. I'll give great attention to the angle of the bridge surfaces, which I want to be dead-level, not angled. I don't want to ask for bridge rotation and that is why.
Next, I'll release the overhead machinist jack preload and measure again, recording my findings. What will at this point have happened is this: the bridge downbearing will reflect a graduated setting that is a direct analogue of how the soundboard likes to flex when all the weight of the strings presses down on it. It will read more than a dime's downbearing in the middle of the soundboard. If I pressed down on the bridge 1/16", the downbearing will in the middle read a dime-plus-1/16". What will have moved less is the low tenor and the bass as well as the high treble.
Now, you may ask, How does this even remotely approach the accuracy of exact feeler gauge measurement coming from angle and trigonometry calculations for 0.5 - 1.5 degrees of string angle on a fully loaded soundboard? Well, it does take those calculations into account. For a 1.5 degree downbearing where a rear string segment is 2" long, the feeler gauge setting is a dime's thickness. Down to the middle and the tenor of a 6' piano where that rear string length varies but is usually more than 2", more distance at the feeler gauge distance is indeed needed as well. But the soundboard also flexes a lot more readily - initially out in the middle area as well. And that is where it was also compressed the most by the machinist's jack. So why did I choose to proceed this way? Here's why. The dried-down board at 5% is set to its lowest crown it will ever be asked to be. More precisely, yes, it could go lower to 4% and yes, the board will still have a lot of crown when that happens, because it is a hybrid precrowned rib, compression-crowned board. It's designed to rest under load with a substantial crowned shape its entire life. So, with the board as low as it will ever be, I proceed to further "flatten" it out in the middle to simulate string load or weight. I then set the bearing to give the board an even bearing setting that is close to final loaded bearing in the low tenor and high treble, and naturally higher in the middle of the board - because it needs to travel farther before arriving at equilibrium under string weight.
A word about the typical cantilevered bass bridge of the smaller 6' grands here. What I like about this technique is that it also introduces accuracy to the setting of the bass bridge bearing. With some preload to the middle of the board - on the long bridge - the bass bridge bearing can then be set up while the board is close to where it will live. And that is a good position for setting the bass bridge bearing setting. This is why: the treble bridge bearing affects the bass bridge more than the bass affects the treble. It has more leverage on the board being closer to the middle. So, pressing the treble bridge down while then setting the bass bridge to "have" a dime's downbearing is pretty much perfect for the small grand.
Now back to the overall mockup. Here's why I chose this route and what will happen afterwards. I chose this route first because I really wanted to do the bridge work at the bench and not in the piano. And I noticed that Steinway NY didn't even notch the bridge inside the piano. I then chose to use a combination of a dry board and preload from overhead, to take some of the guesswork back out of the process. The dry board combined with the overhead pressure virtually guarantees that with a dime's downbearing, the piano will always have a subtle and beautiful amount of downbearing. And then, here's what I not only expect to happen, but what will absolutely happen. The board will be glued into the rim at 5% EMC. It will then climb to 8% EMC, normal indoor figure. When it does this, it will "take on" a downbearing that goes thus: the middle of the piano will balloon past the dime thickness setting, past the dime plus the 1/16" of preload compression from the jacks. It will land somewhere north between 1/16-1/8" above that. This works out because in the middle / tenor, at 1 degree bearing with a 6" rear string segment works out to a dime plus a 16th of downbearing. And what happens in the tenor and midsection of a lot of pianos, is that the rear string segment increases there. And at the same time, we want about 0.5 degrees of bearing, tenor and 1 degree, midsection. What all the variables circle around to with this setup is that we can just use the dime combined with a preload to set our bearing. Yes, I would probably confirm a few things along the way and check my rear string lengths and the exact feeler gauge dimension for given areas of the piano. But the general idea is to actually incorporate the natural resistance of the board by compressing it during the setting of the bearing. In this way, we can actually allow the soundboard to speak for itself and dictate what it needs a bit.
Later, when the piano is strung, here's what I'll expect to see. The soundboard will compress to some amount. I expect it to compress in the middle, roughly the same amount that it swelled from going from 5% to 8% EMC. And once it compresses about that amount under the strings, somewhere between 1/16" to 1/8", it will have arrived at .5 degrees of loaded downbearing bass bridge and tenor section of long bridge, 1 degree middle of long bridge, and 1.5 degree high treble. Notably, I expect the high treble to be nearly constant with respect to downbearing throughout the process, rising and falling only a slight amount and leaving us with a dime's downbearing, 0.053" which comes out to 1.5 degrees with a 2" rear string segment.
The bottom line is that if you think through this whole situation you can find a setup that narrows your window of extremes and even invites the physical materials to speak their mind a little bit and dictate their specific needs. We want the piano to always have some bearing. A dry board pressed down slightly during bearing work gives us that guardrail and promise on the low end. On the high end we know that we are going to allow the piano to pick up a reasonable target bearing under normal humidity conditions. And we allow that under severe humidity, the board will experience some increased compression crown that will result in some excess bearing. But on the other hand, we won't likely ever see negative bearing. And with good instrument care, the bearing will be where we want it to be.
I'd like to give credit to Mario Igrec's book, "Pianos Inside Out" and to Travis' "Guide to Restring" for helpful info, as well as to Google Gemini for aggregating an enormous amount of piano data for reference, and for making a large quantity of mathematical calculations as I checked and rechecked my thinking and intuition on this. Lastly, the folks who first taught me rebuilding, Sam Powell and David Hughes as well as many others.
Lastly, this specific type of soundboard is designed to always have crown and to accommodate humidity extremes with less destruction of the spruce soundboard panel. It is designed to tolerate extreme shrinkage and expansion extremes with less likelihood of cracking yet with continuous crown and downbearing in all conditions. This is modern soundboard design and construction
------------------------------
Tom Wright, RPT
------------------------------