Load cells and weighing multihull builds

I called them and it is $75 a day for a single scale. So renting four would be $325 with taxes. So I opted to buy ine unit at $250.

Just wish I understood the physics of weighing a single corner. Probably going to weigh a single corner, add 200# to each corner and weigh again.

 

This would not provide any proof of the method.
Just take your bathroom scale and a table with 4 legs. Go around and weigh the table partially at each leg and sum it up. Then put the whole table on the scale and compare the results.
Additionally you can do it while placing a certain weight on each corner.
Next addition could be: while one table leg is resting on the scale put something under the opposit leg which will rise it for an inch or so.

 

To weigh each corner independently and get an accurate result, you will need to be able to measure deflection and do a fair amount of math. An easier way is to hold the boat from the outermost corners. Lift the bow and then the stern.

 

Are you able to weight 2 'corners' at the same time?

 

I'd need a 15' beam capable of handling say 3000 pounds on each end or at 12.5' anyway..and it'd not be able to roll over easily... I have some 15' stuff around here, but not metal beams; and the timbers would be pretty impractical...so, not really..neighbor a mile down the road might have something, but hauling them is fun and loading them fun[ner].

 

Fallguy
Is your goal to get the LCG as well as the weight?

 

@fallguy
You said in #24 you did order the live stock scale with 4 sensors. Did you cancel the order? If not, perhaps it is worth to think again over using it.

 

I'll try to make this understandable. I'll use the above mentioned example of a table with 4 legs as the physics is the same.
The table shall be rigid and rectangular and symmetrical in shape and weight. If it stands horizontally on a plane ground its weight is distributed evenly and at ground level each leg will bear a quarter of the tables weight. If you lift one of its legs to put a scale under it, the rigidity of the table will cause the table weight to roughly borne by only two legs. The leg on the scale and the most distant leg at the diagonale corner. As the scale has some thickness, say 2 inch, in an ideal case the 2 neighboring legs are ending in the air 1 inch over the ground. In reality there will be a small imbalance and a third leg will touch the ground but bearing only a tiny part of the tables weight caused by the imbalance.

The weight on the leg at the scale will be less than a half of the table weight (even with no imbalance) because the VCG of the table sits about 3/4 of the table height supposed the legs and the table board are weighing about the same. So the tiny slope of the table top will cause the line of action of the weight originated in the CG (plumb line) not meet the line between the leg on the scale and the opposit leg in the middle (see sketch).



The sketch shows a suggested katamaran on two load cells (LC1 and LC2). The plumb line from VCG(0) would meet the ground line at M(0) in the middle of the distance between LC1 and LC2. If you lift one of the demi hulls, causing a tilting of the kat by an angle "beta", the position of VCG will move to VCG(1) and the plumb line will meet the grount at M(1), which is s(1) apart from M(0). So the LC1 will read more than LC2, supposed the readings before were eaqual.

 

It'd be kinda nice. I expect I am heavy to starboard. Watertanks are 53 gallons to port which would offset captain ad well.

 

Weighing one corner at a time will not be accurate due to deflections. You can lift one end of the boat by making a cable or rope bridle and attaching a post to it. Install the scale under the post and weigh one end of the boat. To keep the post plumb, use guy lines going fore and aft.

 

The lifting must all be done from below. I have the equipment.

I disagree that deflection is an issue. The beams and connections are really strong. This boat lifted on 3 points will walk uphill on the fourth and deflect bearly zero, but you wouldn't know that from details herein.

Using contributor @Heimfried 's drawing; the angle from side to side would be inv tan(2/150) or 0.763 degrees.

This means that if I do nothing to the other corners; they will appreciate a moment of 2" of additional weights port to star and roughly .5 degrees (can be measured) for and aft. I don't know the distance and it is below oF today; so not out there a/o this moment.

Think of a massive structure the size of a football field. And weigh one corner at a time. The amount of the moment becomes less and less as the size grows and ftmp, the scale carries nearly the entire weight and the error approaches zero.

Making some assumptions here. VCG of 5', weights are even at a horrid 2500 pounds per point. Scale is 2", no correction is made, scale to starboard.

Angle beta is 0.763 degrees. The moment distance is 0.0665"(edit for math)x12, some significant figure errors, touche'.

The error is that the port side appreciates 0.0665" x12 of the weight from the other side.

The moments.

The moment from the static port side is 2500•6'3"=187500 inch pounds
Starboard same.
The dynamic port is
2500•(75+(12x0.0665") = 189500 inch pounds
Starboard is 185,500 inch pounds.

The port side appreciates more of the weight. The scale only reads weights, not moments. The port side scale will read 189500/375000 or 51% of the weight. Here, of the 5000 pounds total 2520 pounds and the starboard scale will read 2,480 pounds. The error due to the uneven boat is 20 pounds per side or 20/2500 percent.

When weighing the other side; same thing happens and so the two errors add to 40/2500 pounds.

The same factor exists for and aft, but the moments are less as the distances are greater. The maximum error is about 80/2500 percent. Now, 2500 was a worst case and we can shift the correction firther to best case, say 2000#.

I say doing nothing is acceptable; my errors will be low weights, but amounts can all be adjusted up by 80/2000 percent to be conservative.

corrections welcomed

 

@Heimfried

Back in about 1982, I had a German teacher and she often complimented auf deutsch. Now, I have not remembered th word she used for many years, but after your drawing, even if I am wrong in my follow on analysis. To you, I say thanks and ausgezeichnet.

 

With a post you are doing the lifting from below. There are many ways to skin a cat; some more complex than others. Good luck.

 

@fallguy
I suppose you didn't try it so I did it. Just to verify (or falsify) my thoughts about this.
I took my industrial scale and cleared a table at a working corner in the attic.

Firstly I turned the table (legs up) and put it on the weighbridge. The scale reads 10.80 kg.


Then I liftet one leg and put the scale under it, leaving the table free to set its third leg as it comes. The scale reads 4.72 kg - not exactly a quarter of the 10.80 kg.


After that I put a plastic pot filled with concrete on the table above the leg on the scale. The scale reads 7.16 kg. The pot's weight is 2.70 kg.


Subsequently I put some wood scrap (together same thickness as the weighbridge) under the leg diagonal of the leg on the scale and used my fingers to balance the two free table legs.
This way I could find a position at which the scale reading was nearly the half of the table weight (no pics, because my hands were occupied).

Putting some wood pieces under the 2 (until now) free legs I was able to manage the scale reading zero without a visible gap between the leg and the weighbridge surface.
(What looks like a gap is the shadow of the rounded edges of the legs end.)

 

According to your test, this is not an accurate way of weighing an object. You started with a known weight, then the partial weights vary according to how much you lift each corner.