Draft buffers

[Tim Harrigan]

Carl Russell;17310 wrote:

Why don’t you think that the ox yoke acts like an evener?

I would think that draft angle would have more effect than you seem to indicate.

For sure draft angle is important. A greater draft angle provides more lift and allows the team to get under the load and keep the line of draft through their center of gravity as they lean into the load. I understand competitive pullers often like taller animals such as the Chianina for just that reason. It is not hard to put a pencil to this question and calculate theoretically how draft angle can change the pulling forces. It is a question of the combination of horizontal forces and vertical forces. Over a certain range increasing the vertical lift with a greater draft angle allows the team to carry some of the load and reduce the horizontal pull requirement. One day I decided to test this and see how it actually played out on the ground.

I loaded my stoneboat to 1000 lbs with sand bags. I wanted to check pulling forces with draft angles of 0 degrees, 20 degrees and 30 degrees and also with the load distributed on the boat evenly front to back, front loaded on the front half of the boat, and back loaded on the back half of the boat. Of course that is not possible with a team so I used a tractor with the boat hooked to the front-end loader so I could set the height and hitch angle. After I measured all those I pulled the stoneboat with the Will and Abe shifting the load from front to even to back. The graph shows the results.

With the tractor pulling at 0 degree hitch angle the average forces were greatest and increased as the load shifted from the back (428 lbf) to even (491 lbf) to front (514 lbf). At 20 degrees, the pull in the chain was lower and progressed from 368 lbf back loaded, to 452 lbf evenly distributed to 513 lbf front loaded (same pull as at 0 degrees front loaded). At 30 degrees reaction in the chain was generally lowest progressing from 366 lbf back loaded (same pull as at 20 degrees) to 396 lbf front loaded. So angle of draft clearly made a difference but it depended on how the load was distributed when the tractor was doing the pulling.

When Will and Abe pulled the load the average draft was 400 lbf back loaded, 396 lbf evenly distributed and 400 front loaded. If you compare the even loaded 20 degree angle pull (452 lbf) with Will and Abe at 18 degree angle (396 lbf) you can see that load distribution did not make any difference.

I need to look at this in more detail at some point but it almost seems like Will and Abe were able to make some adjustment in their pulling technique to keep the pulling forces minimized as the load shifted from front to back. Now the tractor pull would clearly be a more steady pull with less pulsing than we see with drafts as they step into the load. That likely has something to do with it, but right now I am hard pressed to offer a convincing explanation. Carl earlier mentioned an intentional energy management on the part of the team, there may be something to that.

Carl, I do think the ox yoke acts like an evener, but I do not think it is as efficient as a horse hitch in buffering the forces that are absorbed by the team. And, hitch angle is important, but less so with a wagon than a log or boat because the angle is much lower because of the higher hitch point with the wagon. Also, the wagon offers less resistance that the log or boat and that has an impact on the dynamics of how the team respond to the load. It is not that I think those things are not important, just that I think they were somewhat minor factors in this situation.

[Andy Carson]

Tim’s experiments show that with a correctly tailored buffer (such as a nylon traces on the wagon) can not only decrease the frequency of high draft forces, but also reduce the average draft. In my mind, it is hard to not see how a reduction in the average draft would not help an animal complete the job. Honestly, I had expected that the buffers would act more to “even out” the loads, redistributing areas of high draft to areas of low draft. Even though this may keep the average draft the same, the system might still be very useful. To give a concrete example, I condition my horse with a sled on a half-hour long trail loop. I usually warm up with an empty trip at 500 pounds (me and the sled), then load up with cinderblocks to give her a workout. If I load up the sled to 1800 pounds, she can make it around once with only a couple short breaks, but she’s pretty pooped when we finish. For the two trips together, that would be an average load of 1150 pounds ((500+1800)/2). Now if I load up 1150 pounds and do not fluctuate the load, she can pull it for a very long time and is not nearly as taxed in the process. This is a pretty extreme example, but these examples demonstrate how two work conditions with the same average draft (but very different maximum loads) can be very different in how taxing they are. So, redistributing the work load such that the power required is as constant as possible can be extremely effective. In theory, one might think that since animals do not have gears and can pull very heavy loads slower, they could learn to regulate their power expenditure so as to not overexert themselves on heavy loads. I am sure this helps, but I am not convinced this is the most efficient way to even out loads if there are other options available. In theory, humans (as we do not have gears either) should also be able to do the same thing, but how good are we at it? I think a bicycle is a great example, anyone who has ridden a bike up a hill in a tall gear knows that it is possible to stand up and push through the hill using slow powerful strokes. Try shifting down and you’ll see you get up that hill much faster with less exertion. The human body is simply more efficient with repeated light loads compared to a few heavy loads, even though it is the same overall amount of weight that must be elevated the same distance. It is interesting that when multiple bicycle speeds were first introduced and promoted around the turn of the century, many established racers said that these multiple speeds are “unnecessary” as a “healthy young man” can adjust to hills by pedaling with slow powerful strokes. Many were convinced by the argument and it was such a controversy that in 1902, The Touring Club de France organized a 200 km race between a professional racer on a single speed bicycle and a young woman on a three speed. You can guess who won. The beaten professional, wrote this after his loss: “I applaud this test, but I still feel that variable gears are only for people over 45. Isn’t it better to triumph by the strength of your muscles than by the artifice of a derailleur? We are getting soft. Come on fellows. Let’s say that the test was a fine demonstration – for our grandparents!” It took another 30 years for multispeed bikes to not be seen as “tools of the weak,” but once allowed, they quickly made jokes of nearly every racing record made on a single speed bike. Now, I doubt the use of a draft buffer will be as dramatic as the use of multiple speeds on bikes, and there are many differences between the two cases, but it is interesting to hear some of the same arguments that were mentioned 100 years ago… I really believe the concept of the draft buffer is sound, I am more concerned about the practicality of the system.

[Tim Harrigan]

I agree that reducing the draft is something we should set as a goal. It is the practicality of it that is a concern. It seems for the most part there would be a bigger impact in most cases with a focus on proper fit and adjustment and conditioning. Draft buffers are more strategic. I will be interested to see what you come up with.

[Tim Harrigan]

This graph shows the pulsing of the pulling forces of a sled as the team steps into the load. This graph showing the speed looks alot like the earlier one showing the pulling forces. The spikes in speed look a lot like the spikes in draft. Read the sled speed from the left axis, seconds on the bottom axis and pull on the right axis.

I changed the right axis on the ox drawn sled graph to the same range as the horse drawn wagon draft for ease of comparison. The horse drawn rubber tire wagon average draft was 492 lbf and the bounce was about 350 to 700 lbf. The average ox drawn sled draft was 476 lbf and the bounce was about 275 to 675. So those look pretty similar even though comparing a wagon to a sled is not the best.

Clearly there are differences in anatomy between horses and cattle. I am not so sure cattle movement is as choppy as you described. Where horses and cattle are similar is most of the power for draft comes from the hind quarters. The slight acceleration and deceleration that causes the pulsing is accentuated under greater resistance. The deceleration is the instance when one hind leg is extended forward ready to push and the opposite hind leg is extended back having just completed the push with that leg. The acceleration is when the forward leg begins pushing through. Both buffer the pulling reaction throughout their body.

Some of the things related to fluidity of movement could be evaluated but just measuring forces is going to leave a lot of unanswered questions. To get at what you suggest you would need video cameras to look at efficiency of movement, pressure sensors at the yoke or harness contact points, probably other things as well. I do not have any plans for that. It would be interesting.

[Carl Russell]

There are a couple of points that need to be made. The animals ability to buffer the draft requirements is not a matter of speed and power, but a matter of the elasticity of their own internal mechanisms, ie, tendons and joints, and how they are employed in the effort to move a load.

Anyone who has worked both horses and oxen can clearly see that the oxen almost stutter-step. They lift the draft with their neck, using their front legs as support, and their rearend as ballast. This very action will cause each step to have to start the load again, in a manner, which will cause a spike in draft each step.

Horses on the other hand, lift with their shoulders by lifting their front end off of the ground by extending their hind legs. Mechanically there is a buffer built into the extension of the triangle between their hind legs and their spine, not to mention the huge rubber band that is the tendon that runs that mechanism. In this way there is a much more fluid movement from one step to the other, and in cases of changing draft demand from the working environment, this mechanism will allow them to physically buffer the effect.

In both cases the more conditioned the animals are, the more effective they are at adapting to the changes. My contention is that the differences in these examples may have as much to do with the differences in how each animal type actually moves in relation to draft.

In this way even the 10 degree difference in draft angle will make a difference in the natural buffer of internal bio-mechanisms, because the lower angle plays directly into the calculation of how those joints expand. If the angle requires that more weight is pulling down on the front-end of the animal then the natural buffer is less effective.

Also, the example that CM used is not really that close to this discussion. Extra weight, or the resulting exertion, is directly related to available power. This discussion is about alleviating the impact of changes in draft. Putting springs into a hitch will not reduce the amount of power needed to move the weight, it will only serve to cushion the impact against the animals as they increase power to the required point. And they will not perform like gears on a bike, as there will be a constant mechanical advantage, or disadvantage, depending on how the teamster has chosen to apply the animal power, ie harness adjustment, conveyance, conditioning.

Carl

[near horse]

The slight acceleration and deceleration that causes the pulsing is accentuated under greater resistance.

Do you think the pulsing is actually accentuated at “the extremes” of resistance – both high and low? I’ve noticed the pulsing when using a light sled that tends to glide forward after the initial pull leaving the team with almost no load, then they “catch up” or take up the slack and it’s another spike (albeit not an overall big load).

It seems that we often mediate the pulsing action, particularly with a heavy load, by allowing or encouraging the horses to accelerate as they pull, somewhat reducing the deceleration over the pull. I’m not as familiar w/ oxen but it doesn’t seem like they are as likely/able/willing to accelerate like I just described. I could be completely wrong about this and in no way am trying to disparage the ability of oxen to pull a load. Just thinking about how we work with our animals.

Would it be possible to factor out all the extraneous effects on draft (harness type, draft angle etc) and look at how these 2 differnet animals move forward under a load? Perhaps using a tread mill and vary the resistance of the conveyor platform ? Could even have 2 smaller treadmills – one for the front legs and one for the rear to isolate the forces applied by each.

Nah! That’s not a good idea.

Putting springs into a hitch will not reduce the amount of power needed to move the weight, it will only serve to cushion the impact against the animals

Carl – exactly. There is no free lunch. It takes a certain amount of energy to move a given mass a given distance. Energy spent stretching a spring will eventually be transmitted to moving the load but at an amount equal to or less than that expended to stretch the spring in the first place. In the bicycle example, if both the professional rider and the young woman weighed the same (as did their bikes) it took the same amount of energy to get them to the top of the hill – she just did it faster. The physiological ability of the human body to recover from this exertion is another story.

[Andy Carson]

Carl and Geoff, I do think you lay out a good argument about how draft buffers would never work in theory and I do appreciate the discussion. If it weren’t for Tim’s carefully laid out experiment demonstrating that buffers (in the form of nylon tugs) actually do reduce the draft required for a job, I probably wouldn’t be as interested in this idea. Look back at the wagon example at the start of this thread, same load and distance, but less draft with a buffer. To me, this directly demonstrates that determining the effort (either purely mechanical of physiological) required to do a job is not simply a matter of determining the average draft and multiplying by the distance. My fluctuating load example, as well as the bicycle example, provide “real world” examples that demonstrate how two different work conditions with the same average draft and distance can be very different in the effort required by the animal or person. I think that also like the multispeed bicycle example, this spring concept is going to need to be demonstrated before anyone seriously conciders the concept due to the Power = Average draft x Velocity argument… I suspect that people are open minded enough to consider this IF it can demonstrate a reduction in average draft (like tim’s nylon tugs) or a reduction in max draft forces. It might fail to do this, but I think it’s worth a test. I’ve ordered a spring and the rest should be pretty easy to put together… We’ll see.

[OldKat]

Tim Harrigan; maybe you said somewhere how you obtained these numbers, but if so I missed it. I suspect that you are using a strain gauge of some sort. Could you briefly outline your methodology in making your measurements?

Thanks,
SRR

[Carl Russell]

Countymouse;17412 wrote:

…. If it weren’t for Tim’s carefully laid out experiment demonstrating that buffers (in the form of nylon tugs) actually do reduce the draft required for a job, …..

I appreciate all the work that Tim put into these studies, but I am trying to point out how these numbers do not support a concept that buffers reduce draft requirement.

If the animals were to stop with each step, then the spikes in draft would be higher, which would raise the average. It is like the Sine-wave analogy to movement. If each movement is softened by the spring then the spikes are lower, reducing the average. If each step is abrupt then each forward movement creates a spike. The average has to go up as the draft reduction cannot go down as there is nothing in place to reduce the resistance against the load.

My point about the differences between how horses and oxen move is that I think this has more to do with the differences in numbers than the different harnessing systems. There are still too many variables between these experiments then I can accept as definitive.

I totally get what you are saying CM, about how we need to experience new concepts before we can truly understand them, but we also need to have comparisons that are uniform in their variables so that we are comparing apples to apples. I would like to see comparisons of the same pair of horses, or oxen, with different harnesses on doing the same work.

Carl

[Tim Harrigan]

OldKat;17414 wrote:

Tim Harrigan; maybe you said somewhere how you obtained these numbers, but if so I missed it. I suspect that you are using a strain gauge of some sort. Could you briefly outline your methodology in making your measurements? SRR

I measured tension in the towing device for horse- and ox-drawn farm utility wagons loaded to 6100 lb. One wagon had 6.00-16 bias ply tires and the other had 5.5 inch wide steel tires. Each wagon was drawn with both a team of horses (16º hitch angle) or a team of oxen (8º hitch angle). The horse hitch consisted of a standard collar harness with stitched, inelastic leather/nylon traces. A North American-style ox yoke with a dropped hitch point was used with a team of Milking Shorthorn steers.

Pulling forces with the standard horse traces or towing chain (ox) were compared to the nylon towropes. The nylon towrope (3/4 inch diameter) consisted of woven nylon strands that formed a hollow shell. Within the shell was a 12 inch long hard rubber core. The nylon towrope stretched 4 inches at a constant rate of 1/2 inch per 225 lbs under an 1800 lb load.

Pulling force measurements were made with a hydraulic pull-meter with a pressure transducer on the discharge side of the cylinder. The pull meter was placed in the towing chain and the pulling forces were recorded at a frequency of 5 Hz with a monitor attached to the transducer. A sub-meter accuracy global positioning system receiver was used to record the position of the implement and match pulling forces with specific locations in the field.

[Tim Harrigan]

Carl Russell;17417 wrote:

I appreciate all the work that Tim put into these studies, but I am trying to point out how these numbers do not support a concept that buffers reduce draft requirement. ……….My point about the differences between how horses and oxen move is that I think this has more to do with the differences in numbers than the different harnessing systems. There are still too many variables between these experiments then I can accept as definitive………………… I would like to see comparisons of the same pair of horses, or oxen, with different harnesses on doing the same work.
Carl

I appreciate Carl’s observation that there are many variables to consider and I agree. We could use more information but we are trying to work through it with the information that we have.

Based on what I have seen I do think draft buffers can reduce draft if we think of draft as the pulling force transmitted to the team. There is a question of draft compared to what? If we compare horses to oxen are we assuming horses are 100% efficient? We can see that the bounce or pulsing in the load from the striding delivery is not much different between horses and oxen. The question is how much pulling force does it take to move a load, and after that, how much force does it take a team of horses or oxen to move the load? Maybe the standard would be a steady pull with a tractor to remove the pulsing of the stride and at a 0 degree hitch angle to remove the lift component. But that takes us again to another system were we do not have the soft tissue animal mass buffering and other things that are going on.

I do not see the pulsing as a negative regarding draft forces. If you look at the stoneboat draft graph with different hitch angles and load shifts you can see that the pulsing pull of Will and Abe at an 18 degree hitch angle was equal to the measured draft with the steady pull of tractor at 30 degrees with a balanced load and much less than a 30 degree tractor hitch with a front load! So is there some draft buffering going on there? Seems like it. The pulsing is not like a full stop and start because you have inertia and momentum with a moving load.

My hypothesis is that the multi-component horse harness with evener and single trees is a more efficient buffer than the ox yoke. I am not seeing much difference in the load pulse from the stride between horses and cattle. The soft body tissue effect is conjecture but it certainly is a good argument for well fitting harness and neck bows. The hitch angle I can deal with by resolving the vertical and horizontal draft components based on the hitch angle. You can do that fairly easily by dividing the measured pull in the chain by the cosine of the hitch angle to compare the horizontal pull of each system. I attached a new graph that shows the horizontal draft components. It looks a lot like the graph of the tension in the chain graph but somewhat lower values because the lift component has been removed. Basically, a greater hitch angle reduces the resulting horizontal value more than a lower hitch angle. The horse hitch was 16 degrees (not 18 degrees, my mistake earlier), the ox hitch was 10 degrees. You still see a lower measured draft with the horse hitch than with the ox hitch.

The horse hitch has multiple components to distribute the load and a relatively soft collar to help absorb the spikes in pull before they are transmitted to the horse. The ox hitch has a rigid beam and a chain with no give. How well the forces in the ox hitch are buffered depends on how the animals move together. If the animals are stepping out of phase with each other meaning one is in the acceleration phase of the pull while the other is in the deceleration phase the yoke beam acts like a class 2 lever where the yoke on the neck of the decelerating ox acts as the fulcrum for the power applied by the accelerating ox with the load in the middle of the beam. That seems like the ideal situation.

Sometimes the animals are in phase where they are stepping forward together and they lose the lever effect. They both push forward and they spike the pulling force because there is no give in the chain and no mechanical advantage from the levering effect. All the buffering is soft tissue and that is not an ideal situation. Soft tissue does not have an infinite ability to buffer forces. That is why buffers for shock loading are important. So in some ways it is just in the numbers like Carl suggests, but the pulling forces are real and perceived by the team. It is true that if you look only at the wagon it should not take more pulling force for the ox than the horse, but the efficiency of the applied pull is different for the horse than the ox. My guess is perception is reality for the teams. If I had to drag one of two loads by myself and I knew one required 100 lbs avg pull and the other required 120 lbs average pull, I am pretty sure which one I would chose.

Carl, you will have to explain your comment about the same pair of animals with the same harnesses for a valid comparison. We used the same team with the same harness, but switched out the traces and they both pulled the rubber and steel tired wagons. Same with the oxen, pulled the same wagons as the horses. Are you suggesting we put the ox yoke on the horses?

Here are the graphs again for convenience, hitch angle/load balance and wagon draft resolved to horizontal forces.

[Tim Harrigan]

near horse;17378 wrote:

Do you think the pulsing is actually accentuated at “the extremes” of resistance – both high and low? I’ve noticed the pulsing when using a light sled that tends to glide forward after the initial pull leaving the team with almost no load, then they “catch up” or take up the slack and it’s another spike (albeit not an overall big load).

Yes, it seems like it. When there is little resistance the load tends to catch up and put slack in the line then snap taught.

[near horse]

In thinking about this more, if we are using the term “buffer” to mean resistant to change, then I can agree that certain devices (like the nylon harness or spring) will buffer the draft on an animal or team.

One other question regarding one of your graphs Tim. Why did using nylon harness on the steel wheeled units result in higher average horizontal draft for both oxen and horses? Didn’t the nylon harness give lower values for the rubber-tired wagons?

[OldKat]

@Tim Harrigan 17418 wrote:

I measured tension in the towing device for horse- and ox-drawn farm utility wagons loaded to 6100 lb. One wagon had 6.00-16 bias ply tires and the other had 5.5 inch wide steel tires. Each wagon was drawn with both a team of horses (16º hitch angle) or a team of oxen (8º hitch angle). The horse hitch consisted of a standard collar harness with stitched, inelastic leather/nylon traces. A North American-style ox yoke with a dropped hitch point was used with a team of Milking Shorthorn steers.

Pulling forces with the standard horse traces or towing chain (ox) were compared to the nylon towropes. The nylon towrope (3/4 inch diameter) consisted of woven nylon strands that formed a hollow shell. Within the shell was a 12 inch long hard rubber core. The nylon towrope stretched 4 inches at a constant rate of 1/2 inch per 225 lbs under an 1800 lb load.

Pulling force measurements were made with a hydraulic pull-meter with a pressure transducer on the discharge side of the cylinder. The pull meter was placed in the towing chain and the pulling forces were recorded at a frequency of 5 Hz with a monitor attached to the transducer. A sub-meter accuracy global positioning system receiver was used to record the position of the implement and match pulling forces with specific locations in the field.

You are certainly methodical and ingenious in obtaining your measurements; my hat is off to you for undertaking this task. I’ll continue to monitor this interesting discussion.

I am reminded what Jan Bonsma, probably one of the most preeminent animal scientists of the immediate past century, said; “Man can’t improve what he can’t measure; MAN MUST MEASURE”

[Carl Russell]

I’m sorry Tim, I had been focusing more on the steel-wheeled wagon looking at oxen vs. horses, and saw too many variables there to just be convince that the difference was related to buffer.

Thank you for posting the comparisons within each team as well. That is interesting, but it still brings me back to where does the energy go that is required to do the work?

The unbuffered must surge forward and reduces the draft more than the buffered one. Somehow it seems there must be a balancing there somewhere.

By comparing the two hitches to the zero of the stationary load, the spikes are not off-set by the reaction forces… or are they, and I can’t see that either?

Can we see these numbers showing high and low spikes above and below the average as zero? (like the wagon on the hay ground chart). That would show me more accurately how the buffer softens the line relating to draft.

Carl

[Author]

Posts