Bow Testing Machine
I need to build an archery bow testing station, At present all the functions are performed individually and manually and are quite time consuming and lacking in accuracy.
The station would basically resemble a crossbow with an automatic drawing mechanism and an arrow would be fired through a chronograph mounted on the front of the machine. The mechanical aspects are not too hard to deal with but I am struggling with some of the electronic features I am looking for.
Features:
A scale that would measure the draw weight at specific or infinite intervals, this should include a hold for highest draw weight and an adjustable audible warning buzzer once target draw weight has been reached
A tape measure that would accurately record the inches drawn and correlate that to the poundage at each designated point. In this case it would likely be about one report per ¼” drawn.
A chronograph mounted on the front of the machine to record arrow speed.
The chronograph, tape measure and scale should all report back to an excel or other program and make some simple computations.
At the start of each test I would have to fill in some blanks to start the test such as arrow weight, and desired draw length.
The computations and report generated should include a graph showing force draw curve, which is just the amount of draw weight present at each ¼” of draw, it should show the amount of stored energy going into the bow, arrow speed and the amount of kinetic energy going into the arrow and then compute the bows efficiency, ratio of stored energy to draw weight, and peak draw weight at specified draw length.
I started on this project before I retired and abandoned it about a year later giving away what I had purchased to someone who said they could complete the project. That didn’t pan out and I am retired now and want to start over.
My first question is what type of person would be the best to contact before I start buying electronics, or better put where would be the best place to look for the person I need. If it is too involved and too expensive I will just abandon before I spend anymore money. If said person tells me he can write the program I need for maybe $2,000 and steer me to the right electronics it still might be worth me going forward and completing it. My hope is that once I have a prototype I can build future models for about $500.00 and possibly sell a few and offset my expense on the first one as well as provide a good testing machine to my archery community.
If you use a linear actuator for the draw you won’t need a tape measure, you’ll be able to control the length of the draw in correspondence with draw weight. A strain gauge can be used to measure the draw weight very accurately, that is if the bow is properly mounted. Look at some Mythbuster episodes and the devices Grant Imahara makes for things like this. He might have done something similar already. Some optical sensors may be all you need to measure arrow speed. The programming may not be the most difficult part, it sounds like something you need a good engineer/builder for to make sure it works properly and measures accurately. And don’t try to design your final product from the get go, first get something mechanically working and evaluate how well it fits your needs. On this board Mangetout may be one of the guys you want to talk to.
OK, IMO the hard part is the sensors and controls – not saying it’s impossible, just the hardest part. Once you’ve got the sensors accurately reading and sending their output to a computer, and the computer controlling the draw, the software from there is relatively easy (and the hard parts are related to exactly how the computer is telling the motors what to do, etc.). The computations could even just be done in Excel or something once all the data is read. And the mechanical seems very straightforward (though you might want to get an engineer to double check things, as I assume there will be a fair amount of stored energy in the bow that could wreak havoc if something failed mechanically).
So what you need is to find someone who has experience with computerized sensors and computerized mechanical controls. Then ask them if this is something they could do and how expensive it would be. I assume the chronograph is pretty much already a complete unit – worst case it’s not computerized and you just manually type the result into your spreadsheet, which shouldn’t be a big deal. Rigging a sensor to see how far the bow has been pulled should also be easy, and getting a computer controlled motor likewise. I’m not sure how hard it would be getting a tension sensor between the motor and the bowstring would be – that’s probably the question to ask someone to see if they can do this.
I do a lot of bow testing and experimental work on bows. I also do a bit of writing on various designs and tecniques an how they affect performance. Happy to report I established 2 new world records this year for distance shooting.
Good point about the chrono, I could just type that in. I could even manually draw it back for that matter. If I used a spool that made less than one revolution and the sensor could count what portion of the revolution I was at and report that continuously that might work.
Back in the old days, we wouldn’t have a linear actuator or digital strain gauge, so we’d roll our own, and need only one part, not two. The main part would be a linear motor (or a rotating motor with a capstan) whose force we could control (either by voltage, or by duty-cycling). We’d first calibrate it to measure how much force we’re applying (for a given voltage or duty-cycle). We’d need a location sensor – that’s the trickiest part I think.
Then we’d use something like a PID controller algorithm to control the location. We’d draw back in steps and hold for a moment and then record the force. Using a PID controller, you might not be able to do a continuous draw, since the force applied would vary due to the simple control law. (That is, it would oscillate and damp out, and you can’t measure the force until the oscillations damp out.)
These days, I believe you can get most of the parts with built-in controllers, but without access to the internals of the controller, so you’d probably need both a linear actuator and a strain gauge. Of course, the linear actuator would have the draw distance measurement built in. The trick is finding one that handles the distance and force. I assume you’re talking in the 100 lb range, for force – maybe more, for some compound bows with high early peaks.
If a pause was needed if I could set it to pause at every inch to compare weight to draw that would be sufficient. Primitive or all wood bows do not like to be held at full draw for more than a couple of seconds as they tend to take set or explode. Bow explosions when testing are not an uncommon experience.
Would the draw on the motor pulling back the bow be an accurate measure of draw weight, and would it loose calibration as the motor got older?
A screw driven linear acuator will count revolutions of the screw so the motor performance won’t be a factor unless it can’t provide enough force to actuate anymore.
What I was reffering to was the amperage draw on the motor changing as the draw weight of the bow increased while drawing it back, would monitoring the amperage draw be an accurate way to monitor weight as it drew back the bow?
Monitoring the current wouldn’t be that accurate and it would change as the motor ages. A screw-driven linear actuator with a strain gauge is a much better way to go.
First, you wouldn’t monitor the current. You’d monitor whatever signal you’re using to control the force, and convert that into force. While motor characteristics can change as the motor ages, a simple calibration procedure would be sufficient.
That said, it’s likely that a screw drive is probably the best bet here, allowing a fairly weak motor to impart a rather strong force, yet a screw drive would not be the best fit for my suggestion (as I said, an old-school approach). The reason is that there would probably be enough friction in the screw drive to affect the accuracy of a force measurement.
So I agree that in practice, a stress meter would be better.
The simple way to avoid breaking a bow would be to draw out to a distance, hold for long enough to measure, then return. Draw out for the next distance, etc. Most likely, I’d use something in between “all in one draw” and “one measurement per draw”. That is, I’d probably measure the first 6 inches with 6 pauses in one draw, then 5 more inches with 5 pauses in the next draw, and so on, or some schedule that minimizes the time spent at maximum draws but gets the job done quickly enough. How many thousand bows are we measuring?
The biggest disadvantage of using today’s whiz-bang devices is that you don’t have the joy of learning how to code a simple PID (proportional-integral-differential) controller. That’s one of the niftiest little tricks for a newbie real-time programmer (along with a low-pass filter, even easier and super useful).
If anyone wants to know about these things, just ask in a new thread. Probably not needed these days. Spoiled youths! (PS: no doubt there are great treatments on the web. But these things are pretty easy to describe and work in rather intuitive ways, so a forum thread wouldn’t be a bad way to do it.)
That does sound like something that might be very useful for what I am doing.
As for drawing back, I wouldn’t have any problem with doing this manually with a crank as long as I had a clicker or something built into the drum that would allow me to pause for a reading each 1" of draw. A manual draw might even be preferable as often an arrow might come dislodged or whatever. Also it would significantly reduce cost, weight, and power requirements. I was hoping to avoid the pause for a reading but that may not be possible if I am looking for accuracy.
There is a ready made tester almost exactly what you are looking for, but it is probably way out of the price range (Tinius Olsen).
Something that would be much less expensive would probably be a Click series PLC (Logic Controller) with some analog I/O, perhaps a small graphical interface terminal, a linear motor, strain guage, and amplifier.
The nice thing about the CLICK PLC is that is is very functional, and yet very inexpensive. I think even the programming software is free.
If you want Ethernet connectivity, there are higher end PLC’s like the DOMORE that are a bit more expensive, but still fully functional.
Just about everything you are looking for can be found at Automation Direct.
Tinius Olsen does have some very cool testing machinery. It would be ideal for testing material properties in only the way a bow maker is interested in them. When bow makers sit around having a few beers and start to dream we often talk about creating a very elaborate facility that could test and answer all the questions we haven’t really pinned down. If about 200 guys pulled their money it might be feasable.
That’s implying he can write & test the program in something like 25 hours - unlikely.
Unless you can sell a large quantity and produce them in a small factory (with consequent economies of scale), this figure seems seriously optimistic.
As an alternative to linear actuator and strain gauge, a pneumatic or hydraulic cylinder might work - draw force will be proportional to the pressure in the cylinder.
