Home-Made Corner Weight Scales
Many years ago I went to TAFE with a friend of mine who we all knew as “Dangerous Dave”. I struggled in my Statics subjects, whereas Dave was a gun at it. He quickly worked out not only how to solve the problems in class, but how to apply the methods in real world situations.
Dangerous Dave concocted a rig to measure corner weights. While not exciting in itself, (there are after all many examples of homemade load cell systems on the internet) what made Dangerous Dave’s design so interesting to me was its simple elegance and an in-your-face use of science.
Dave’s take on corner weight scales was to take four cheap bathroom scales and work out how long the lever needed to be to allow the scales to operate within their functional range.
This turned out to be very simple. On a beam 1000mm long with a point load at 800mm and supported at each end will present a ratio of 4:1. That is, 600Kg at 800mm will cause a force of 120Kg at the long end and 480Kg at the short end. What that means for us is the use of a length of wood or metal on a pivot at one end, the weight of the cars wheel at the 800mm point will be easily measured with a 0-120kg bathroom scale at the other.
Take the measurement presented on the bathroom scales and multiply it by 5. This gives the true mass on the beam.
This can be scaled up for heavier cars, but the ratio does change. A 1200mm long beam with the load at the 1000mm point will require a ratio of roughly 6:1. That is, 100Kg on the scales will reveal a 600Kg point load at the 1m mark. It’s important with spring type scales to keep the weight at around the mid point of the total range as this is where they are at their most accurate.
I quickly jammed together four 1200mm beams made from surplus bathroom wall, and bent a couple of nails over and into some copper pipe as the pivot point. A quick trip to a generic department store and I arrived home with four bathroom scales. There are electronic bathroom scales with a resolution of 100g for $20 each which would have been ideal, but I cheaped out and bought the $8 dial gauge type. I did a calibration test with my own generous bulk, which proved the concept. It was accurate to within a kilogram if paying careful attention to where the needle sits. 1kg on the dial is 6kg of true mass so to be accurate is nearly impossible with these cheap analogue scales but a good approximation is feasible.
With the beam on the scales and zeroed then my Porsche 911 sat on it; it showed a weight of around 40kg each side for the front and about 55kg each side on the rear. Multiplying each corner weight by six reveals the true weight give or take a few kilos. Considering I have the windows removed at the moment and a few other things out of the car, the total weight of 1114kg is probably about right. My main mistake was in not getting the beam level as the pivot point was not the same height as the scales. The revised (and better made) design will remedy this.
This correlates for the odd weight distribution of this car (taking into account the full tank of fuel but missing window glass). To do this properly, the beams and floor must be completely level and all four wheels of the car at exactly the same height, so if you decide to only use one scale, the car must be placed on chocks that replicate the measured wheel height exactly. Even tiny differences in height between corners will distort the reading enough to be noticeable on this setup.
Is it consistent? Yes, it is. Is it able to measure a quarter of a kilogram difference from one side of the car to the other? Well, no; the resolution is quite poor with an accuracy of about +/- 6kg.
Better quality scales or an investment in a 0-100kg load cell will greatly improve the accuracy, but for $20 of materials, it’s a good way to get a thumb-and-squint measurement to help with initial setup. I highly recommend making the frames from better materials than I used. I hate maths as it is an excellent opportunity for a numpty like me to make a fool of himself, so let us know if you gave this a try, and how well it worked for you!