[Andy Carson]

It makes me really wonder. Maybe I am “reinventing the wheel” here and the Europeans already have a system to do this… The picture doesn’t look like they were pulling a wagon with this set-up, there is no breeching and it looks like the horses are to be attached to whatever they are pulling with a dragging chain. I would have guessed this to be a heavy dragging load, maybe a log? I am really curious because when I had first calculated the size of the spring, it seemed that a very large spring was needed. A large spring would be needed to equilibrate and store energy over a 0.2 second interval with the variations in speed that Tim had reported. After having done several runs, though, I can see the spring rarely compresses more than 1 inch over the preset. My interpretation is that with the spring equalizing and buffering the pull, variations in speed are minimized. If variations in speed are minimized, not only would that serve to reduce the spike forces, but would also reduce spring movement significacntly. Long story short, it seems the spring might not need to be as long as I had initially thought. Also, after having bagged on (somewhat) progressive springs, I am beginning to think that they might be better for this application. A progressive spring might make this system smaller and easier to handle, adjust, and implement, and increase the range of loads that could be effectively buffered. It would be very helpful to discuss how the springs that Jason posted work, and what they are effective at doing. Maybe one of our European friends can help us out here.

[Andy Carson]

It is a very interesting design for a buffer using a much smaller and more managable spring. From the looks of it, I would guess these are more for shock protection, but I am not sure. It would be nice if anyone could report on these. They might be able to tell us if the springs are compressed very often during regular use (as opposed to only shock loads) and if they seem to help for other applications, such as starting a heavy load.

[Andy Carson]

@Tim Harrigan 17743 wrote:

…Proper harness or yoke fit and design, hitching, conditioning and reasonable expectations have potentially a much bigger impact for our community of interest than a draft buffer…

Having bought an out of shape horse, I can report that huge increases in sustained pulling power are possible with steady work and simple conditioning. Those increases make the 20% increase possible with the buffers look like a joke, no doubt. Still, an animal that is in good condition that is working near it’s limit would have to appreciate a 20% reduction in their workload. From the experiments I’ve done, it does seem like a particular buffer will have a somewhat narrow range of average drafts they will buffer at any particular setting. This isn’t a serious limitation in my mind, because I think there is only a small range of tasks that by their nature, would benefit from a buffer. For a small increase in efficiency to have practical utility, I would think the task would need to be long term. Similarly, I doubt that there is an practical utility in bothering with buffering really easy jobs. That’s what led me to pick a “useful range” of 300-500 pounds for one horse, and the buffer seems to require little adjustment within this range. Tim is right that since this is not near the short term limits, it has been difficult to demonstrate effectiveness. In giving directions for springs and whatnot, I am not advocating that everyone use the buffer for everything, I was simply saying “Hey, here is a simple tool that is pretty easy to build. Give it a try. Use your judgment and common sense, and see if it helps out.” After several runs with different set ups, I am becoming more confident that there are benefits of this buffer when adjusted properly. Still, only when a several people try it out at different tasks can we reach some sort of consensus.

[Andy Carson]

John,
I think this is a good point, and yes, there seems to be a different “feel” to a load (at least when I’ve pulled with a buffer). Ar first, my horse seemed to think that since she was not feeling “hard bumps” when she pulled (and especially started) a load, the job was going to be easy. The sled was still heavy though and was hard work to pull… So even though she wanted to take off at the beginning, she learned what the “feel” meant after a short time (maybe 5-10 minutes) and learned how to pull with it. I think that animals have great intelligence when it comes to making physicaly adjustments to different physical “feels” or “rhythms” but I see your point about the “insulating” effect. I am not entirely sure how I would know if the buffers were inhibiting the animals reaction rate. On the runs I’ve done, there was nothing obvious, but I wasn’t looking for this effect. What kinds of behaviors should I be watching for?

[Andy Carson]

This is very cool! I’ll do a little math for John, since he seems to find it so entertaining. If each of the 14 drafts is pulling 250 lbs (13-17% of body weight), than that’s a total 3500 pounds of force (250*14). As it’s acting on a 12 foot lever arm, that’s 12*3500 = 42000 lbft of torque. That’s about the same as the torque generated by about 65 6.7L Turbo Diesels running at maximum torque (listed as 650 lb-ft max). Of course these torque numbers are kinda distracting without taking into account the radically different rpm’s that the horses and the diesels run at, but it is still fun to see such a number. Cheers!

[Andy Carson]

Geoff,
I am glad you are interested in trying out a spring buffer too. The buffer is really pretty easy to make once you have the spring. If you are going to get one for a team, you may have the best luck with a 12 inch (more or less) long spring with a rate of 250-300 lbs/inch. I think mine (119 lbs/inch) could use a little more “umph” and of course a team would need twice that. McMaster-Carr, sells a 12 inch long spring with a rate slightly over 300 lbs (96485K455) for $40, but you might get off cheaper by finding one at the junk yard. This is in the range of coil springs used for the rear suspension on some passenger cars, but there is alot of variation in these cars. Alot of these springs are going to be “progressive” and will be easy to compress at the beginning and become more and more difficult as the spring is compressed. The wires coils of progressive springs are usually close together at one end and wider spaced at the other end. I thought it was easier to do the modelling on a nonprogressive spring, but a progressive spring might work too. You can roughly determine the spring constant on a nonprogressive spring, with some weights and a measuring tape. You would be looking for a compression of about 1/4 of an inch with a 60-75 pound weight (or 1/2 inch movement with 125-150 pound weight). A progressive spring would need to be calibrated thoughout it’s load range, which would require a more complex set up, but might be the best way to go. Still alot to know…

By the way, I took a look at the paper that you posted. I was very interesting and well written. I thought the measurement of power versus speed was very interesting, as when the horse slows from 2 m/s (4.5 MPH) to 1.4 m/s (3.1 MPH), the power drops off only slightly, but the power is cut in half when the speed is reduced to 0.5 m/s (1.1 MPH). These lower speed higher draft numbers were from another paper they referances and are the white squares in figures 6 and 7. I was also entertained by the comment that the draft force was set at an upper limit of 20% because the Thoughbreds “became agitated if asked to exert greater force.”

[Andy Carson]

Mitch, I’m glad you had a chance to try out a spring too. Even those many of the observations that both you are I am making are subjective, if different users see the same things, then they are more likely to be true. I do think that the springs in a spring tine harrow would act as buffers, but I am not sure if they would be in a range to be effective. It seems that the spring constants and response rates need tuned to a degree for the buffering to be most effective, and I just don’t know if the spring tines are there. I think that a committed buffer designed to respond to the predicted loads would be ideal, and yes, Geoff, I think that a spring tooth is great application of this technology. So comparing Mitch’s and my runs. Mitch, correct me if I’m wrong, but I think you are using a spring without a preload? Some differences might be attributed to this.

1. We both see a reduction in the power required by the front limbs
2. Mitch noticed a greater tendancy to walk, I can see this without a preload, but with a preload, my horse has a tendancy to go faster
3. I noticed a swinging action of the sled without a preload, which alternated high and low drafts and annoyed my horse. Mitch reports a jerky motion that his horses did not seem to be annoyed by. I am not sure if this refers to the whole implement or individual springs, but if it is the whole implement, than perhaps the “jerkiness” vs “swinging” is the same phenomenon occurring as a different frequency because of different spring constants and spring lengths.
4. I noticed a definite smoothing of forward speed, and I am not sure if Mitch sees this. I have to say, though, a periodically nosediving springtooth would be a lot more difficult to smooth than a sled.
5. I noticed an easing of effort to start loads, I am not sure if Mitch sees this…
6. Mitch reports as overall easing of effort, and I do not see much of a difference in effort. This is likely because my horse is a workaholic and if she feels like the load is easier than she just goes faster. I do notice that my horse speeds up with the buffer so perhaps these observations agree as well.

Very interesting to compare experiences. Perhaps if we could hear a little more about Mitches spring it would be helpful. What are the dimensions of the spring? Do you have an idea of it’s constant? How did you hook it up? Does it seem to be compressing or extending very often? How far does it compress when it is in action? Does it ever seem to bottom out?

[Andy Carson]

You were probably going to do this already, but I would be curious to know what would happen if you compared a spring the spring to the “no spring” with the same number of horses and the same draft angle. In the best case scenario, we are probably chasing a 20% improvement in overall efficiency. This is based on Tim’s draft force and speed vs time chart where there were a total of 13 0.2 second peaks over 8 seconds total measurement time. That means the animal is producing high peaks 30% of the time. According to some math I did previously in the thread, these peaks may contain up to 67% wasted energy. So about 70% of the forces would be uneffected and the left over 30% might be reduced by 2/3. That results in a 20% reduction, but it assumes a lot and I would be suprized if this level is achieved.

[Andy Carson]

With a 4 inch preset and a 920 lb load, I see action when the terraign is moderate to difficult and the horse is in the “powerstroke.” This action produces a audible, steady beat for most of the time we are going. It sounds like “ka-chunk, pause, ka-chunk, pause, ka-chunk, etc.” The “Ka” is the rubbing of the threads on the metal sprign holder as the spring is compressed, and the “chunk” is the sound of the set nut coming to rest on the holder when the pull falls below the preset. So, in one way, the spring is in action most of the time (as it only fails to periodically compress if I am going downhill). In another way, it is not in action very often because when it compresses during the normal stride, it only does so for a fraction of a second and quickly falls back to the preset point until the next stride. I think this is line with the periodic peaks in draft force that Tim has measured. I had thought the average draft on this load might be 368 pounds (920*0.4), which would only decompress the spring when the load is exceeded by 108 pounds ((4*119)-368), a nearly 30% increase. Perhaps because the trail is muddy in parts and full of ruts the draft is not best estimated by simply multiplying the weight by 0.4 it seems substantially higher… My thought was that reguardless of the terraign, I can vary to load, observe the behavior of the spring and the horse, and later “zoom in” on load and spring ranges that seem to warrant further observation. Using this stratagee, I can see “interesting behaviors” at all spring/load combinations where the spring is compressed briefly during the “powerstroke” and returns to the preset point during the rest of the gait. In other words, “Ka-chunk, pause, Ka-chunk, pause” seems interesting. “Kaaaaaa-Ka-Kaaaaa” does not seem interesting, and no sound does not seem interesting either. It is a good that the “Ka-chunk, pause” type of setting is pretty flexible and can be achieved over a wider range of spring/load combinations. Also, this observation gives the user an easy way to adjust the preload on the spring without requiring precise calibration. Interestingly, it seems that the lower preload settings (Kaaaaaa-chunk) seems more comortable than the higher preload settings (K-Chunk). What seems most uncomfortable is if there is no “Ka” or no “chunk.”

[Andy Carson]

Another report from the spring draft buffer… In an attempt to find some more dramatic differances to display, I set the preset on the spring to 5 inches (595 lbs), loaded up the sled to 1130 lbs, and pulled some hills. My thought process was that by increasing the draft forces and efforts the point that they are much more difficult, differances in the buffered versus unbuffered system might be more apparent. I made a number of interesting observations, some seem obvious now but might be interesting to discuss nonetheloess…

1. When pulling a heavier load, the horse is more likely to bob it’s head, even if the system is buffered
2. The buffering capacity of this spring is not high enough for this particular load, especially on hills, where it bottoms out every now and then
3. Starting the load using the spring buffer appears very easy and the horse simply walks off without any double leg thrusts, even up hills
4. The buffer maintains a steady forward speed even when the horse is moving very slowly up a hill, and as the load never completely stops, the horse is more comfortable pulling at a very slow speed
5. Some of these hills make my horse very tired and I am used to seeing a little shaking in the shoulders when I give her a break at the top of one particular hill. The use of the buffer completely eliminated this, maybe because the horse didn’t have to rely on a strong (though brief) front limb push to keep the momentum going and allowed the rear legs to produce a larger percentage of the power…
6. I saw no obvious improvement in overall efficiency, in other words, a the end of the run, my horse seemed just as tired with the buffer as without. There are several explanations for this, but it still remains that the buffer is not clearly demonstrating an advantage in overall efficiency, at least by a very subjective measure…

After this set of observations, I switched back to the 4 inch preset (476 lb) and 920 lb sled to verify if what I had observed previously was reproducible. The reduction in vertical head movement is reproducible when the spring is in action. When the spring is not compressing at all (as in a downhill), the horse moves normally, likewise, if the spring does not return to the preset point (as in an uphill) the horse moved normally too. This different action on the part of the horse occures when the spring is rythmically compressed at a specific range of loads. It’s kinda cool to watch, almost like a dancer finding the beat, and there were several parts of the pull when this rythum could be maintained for a while, but I am beginning to think that this type of fine tuning might not be the best way to go. Another couple cinder blocks of load, for example, and the “beat” could never be found again…

I think I am going to focus on determining if and demonstrating that the buffer
1. Eases the starting of a load
2. Smoothes the forward speed and prevents stopping (and the energy expenditude to start again)
3. Reduces maximum draft forces
4. Eases the strain on the front quarters and allows the hind legs to contribute a higher percentage of the draft power

[Andy Carson]

I did a quick math check and have come to a very different interpretion of the radical verticle movement during galloping which seems to be efficient at storing energy and how that relates to energy storage with a horse pulling a heavy load. From Tim’s force data, the draft pulling a heavy load produces high draft peaks of about 200 lbf (890 newtons) if this force acts on a 1760 lb horse (800 kg), then the accelleration would be 1.11 M/s/s (F=MA: 890=800*A). As the acceleration of gravity is 9.81 m/s/s, this force wouldn’t even get the horse off the ground, much less elevate it 4-6 inches. This means that elevating a smaller portion of the body (such as the head) could indeed store energy efficiently. I think that in the case of the galloping horse, there is a rythmic up and down motion that is efficiently returned and added to slightly with every thrust. I am thinking of jumping on a trampoline, where each strong trust from your legs is extended in action by the springs of the trampoline and returned efficiently upon landing. Just like the trampoline (or a pendulum), there seems to be a small range of rythyms and speeds with which these different gaits can return energy efficiently. Perhaps this is why the animals in Tim’s experiments (and perhaps in real life) tend to top out at a specific speed when under heavy draft. Perhaps this is where the energy return is most efficient. On the surface, it seems that if the work is twice as heavy than the horse could expend the same amount of energy but do the work twice as slow. Maybe it’s just me, but my horse tends to gravitate towards a few different apparent speeds where she can get into a rhythum of raising and lowering her head in time with her foot falls. She seems much more happy at these specific speeds than if I speed her up or slow her down 20%…

[Andy Carson]

Thanks for the words of encouragement. I am certainly not an expert, but I find these types of discussions very fun and am glad others find them interesting as well. This site has really been an incredible source of both practical information and inspiration and I am honored to be able to contribute to the community a little in a meaningful way. This is a field where the “art” probably matters more than the “science”, but a little science can’t hurt. I remember Carl’s thread about how he moved the large oak logs up a steep hill and how he used simple, but cleverly designed stratagees to give the animals advantage and make the job more efficient for the horses. Perhaps with a more complete understanding of the efficiencies and inefficiencies of animals, we, as a community, can find clever ways to overcome limitations and improve the capacity of the animals to do work.

[Andy Carson]

I reread that paper and do not see the resting activity being a serious limitation of this work… Geoff, perhaps you noticed something I didn’t see… I also learned a couple more things of relevant to this discussion that I hadn’t noticed on the first read. In thier introduction, the authors mention that tendons and ligaments can indeed store energy to produce a “bounce.” Later in the paper, they measure the large amounts of energy stored and released by the verticle displacement of the entire animal. This was substantial at the trot and gallop, but negligable at the walk. This is interesting in the context of a horse raising and lowering it’s head a couple inches during a heavily loaded walk. When a horse in the study was producing alot of power (as in a gallop) a verticle displacement of 10-15 cm (or 4-6 inches) was needed to store all that energy. If only half of the body was raised, the body would have to be rising and falling a full 8-12 inches… Different horses, different loads, and very different speeds, but it seems unlikely that raising and lowering the head slightly would fully recover the energy.

[Andy Carson]

Sorry Geoff, it’s in the Journal of Experimental Biology, Vol 202, Issue 17, pages 2329-2338. I was refering to figure 7. The weakness of this (for our purposes) is that the horse is not pulling a load. I suspect without a load, a horse has very little interest in conserving energy and it might be unlikely that they would be motivated to learn… I also found it interesting that the more suspended the gait is, the more energy could be stored and returned. A gallop stores and returns about 90% of the energy, but that requires very drastic movements on the part of the horse and doesn’t really have that much to do with our discussion…

[Andy Carson]

H. Ortiz-Laurel and P.A. Cowell. “Power Output Measurement on Draught Horses”. Agricultural Engineering International: the CIGR Ejournal. Manuscript PM 07 001. Vol. IX. August, 2007.

Another fascinating read where the authors have studied draft forces on a 900 lb highland pony pulling various loads. Althought they tested a maximum pull of only 160 lbf, thier system allows them to measure both the verticle and the horizontal component of the developed force on each individual limb.

Some interesting conclusions from this study:

1. The forelegs of this horse carried greater weight at all draft loads, but the load was progressively transferred to the hind legs as draft increased

2. At low draft loads, the forelegs and the hind legs made nearly equal contributions to the propulsion force. At a draft load of 160 lbf the hind legs contributed 57% percent of the horizontal pull. Although they didn’t test higher loads, the curve of power vs load generated by the hind legs looks very linear and is a long way from any plateau. The power vs load for the front legs, though, seems to reach a plateau near the maximum tested.

3. The difference between the power developed by the hind legs and the power developed by the front legs grows as the load is increased.

Interestingly, the authors also indicate that the shoulders of draft horses contain 4.5% of the total muscle, while those of race horses contain only 3.6%. I think this indicates that the front limbs are very important for something, even though they seem (on the surface) to be less important as draft increases. I think this supports the idea that while the majority of the power generated by horses under heavy loads is generated by the hind legs, the load that can be maintained at a normal walk (rather than with periodic double hind leg thrusts) is limited by the ability of the front limbs to maintain foward momentum during the time between single hind leg thrusts. That makes me think… Tim, on the graphs you generated for starting a load, were the horses starting with a “double leg thrust” or simply “walking off”?

[Author]

Posts