Ox logging and ground skidding draft measurement video

[Tim Harrigan]

I had a great time in the woods this weekend ground skidding some logs with Will and Abe. I took some video and tested my pull meter and data logger recording pulling forces at 5 measurements per second. This video includes a comparison of pulling forces or a 1,050 lb log using a stone boat, logging tongs and a choker chain. Click on the following link to view the video.

[video=youtube_share;ou0Tav-jzfs]http://youtu.be/ou0Tav-jzfs[/video]

[Carl Russell]

Dude:cool:, THAT’S what I’m talking about…

Nice job Tim.

I really think this is the next wave of DAPNet using the web to advance our sharing.

Thanks again Tim.

Carl

[dominiquer60]

Great job Tim.
It is nice to see that the method of least force is also the better way to keep a saw log clean.
Erika

[Anne]

Hallo Tim,

Great work!! In the woods as well as in front of the computer!!

Thanks a lot!

Anne

[Andy Carson]

One of the interesting things about this group of measurements is that as all three of these twitching methods are applied to the front end of the log, one could assume with some confidence that the drag on the rear of the log remains constant. This allows one to tease out the relative contributions of the front and rear of the log towards total draft. Perhaps this will be interesting…

One could model the total drag on the log as the weight of the front end*friction+weight of rear end*friction. Previous work from Tim shows that stoneboat friction is approximately 0.4x weight. I think this is a good approximation of friction at the rear of the log, which can slide over ground without digging in and tilling it. For this particular log, this would be 1050/2*0.4 or 210 lbf (I am assuming the rear and front halves are approximately the same weight). With a sled under the front end of the log, the average draft is 392 lbf, with only 182 lbf coming from the front end (392-210). One might be surprised that the draft of the front end here is lighter than the rear. I think this is totally due to the angle of draft. The vertical component of a 10 degree angle of draft with a total force of 392 lbf is about 68 lbf (sin10 degrees*392=68). The vertical component would reduce the total applied weight of the front end of the log from 525 lb (1050/2) to 457 (525-68). 457*0.4=183 lbf, which is damn close to 182 lbf.

So, the drag applied by the rear end of the log in this case is about 210 lbf. As the front of the log does not physically lift of the ground in any of these pulls, and the geometry of the rear of the log remains constant, the drag applied by the rear of the log would be the same (210 lbf) in these different situations. I am going to assume the angle of draft also remains constant in these situations. That means the applied force of the front end is sin10degrees*total draft in all cases (which is 92 and 97 lbf for the tongs and chain respectively).

Here’s where the differences really stand out. The average force of the log skidded with tongs is 530 lbf and the chain is 546. That means the drag from the front end of the log is 320 lbf and 336 lbf, respectively. These represent friction coefficients of 0.74 (320/((525-92)) for the tongs and a whopping 0.78 (336/(525-97)) for the chain, which is almost double that of the sled or the rear of the log. Pretty dramatic difference when you dig into the numbers.

I think this is also a great illustration of why getting lift on the front end of a log is important. Actually, these numbers give is a way to determine the optimal amount of lift to apply to this particular log in these different situations. To do this I determined the upward vector of the total draft at 10 degrees, as I did previously. This gives the downward force at the front of the log (525-sin10degrees) and horizontal component of the vector that must be exerted to overcome drag in the front end which is equal to the downward force times the friction coefficient (0.4 for sled, 0.74 for tongs, and 0.78 for chain). The horizontal component of the rear of the log we already know will always be 210 lbf, so we add the horizontal frorce vectors together and use trigonometry to determine the total force vector. One can then use excell to plot all required draft forces at these different angles and get an idea of the ideal angle of draft for these different situations (I have attached the plots below). Interestingly, the optimal angles are not identical. One can tell from the plot below that a sled isn’t very sensitive to draft angle, but is the angle is optimal at about 20-25 degrees. The tongs and choker chain are much more sensitive to draft angle with an optimal angle of around 35-40 degrees. As this angle is probably not achievable with standard hitching methods, the recommendation would be to hitch as short as possible if you are using a chain or tongs. Or you could simply use a sled and not have to worry about angles and such. I think we already knew this, but it’s nice to see the math too.

Interesting stuff, and fun to think about! 🙂

PS. This post contains substantial edits from my previous post due to a few math errors… These were actually pretty important math errors and they changed my overall conclusions.

[Tim Harrigan]

Andy, let me point out a few things about these skidding methods that complicate the analysis. The stoneboat is 8 feet long and weights 175 lbs so the weight of the load was 1225 lbs rather than 1050. The log was 12 feet long so it extended about 4 feet past the back end of the boat, but the log was carried off the ground in back so your front/rear assumption about the friction caused by the rear of the log is problematic. I have a little bit of a problem separating the front/rear contribution to the friction of the log. There would be some lift on the front that would reduce friction as you mentioned but with the tongs and choker chain there would also be load transfer from the front to the rear of the log that would tend to increase friction at the rear. An advantage of the tongs is that they provide a fairly constant hitch point even as the log rolls. The choker chain is a quick and sure hitch, but the chain wrapped perpendicular to the direction of travel acts as a brake, similar to bridle chains on a sled. Also, and what annoys me most about a choker is that the logs tends to roll so the hitch point is near the top of the log, nosing the leading bottom edge into the ground, and reducing the hitch angle by raising the hitch point. These are the things that I think account for the higher average draft and the higher maximum draft for the choker compared to the tongs.

At any rate, your analysis is pretty close but your approach is a little different than mine. First of all, I want to calculate the kinetic friction coefficient without the effect of the lift at the front of the log or boat. Because I measured tension in the chain or the resultant force Fr, at perhaps a 15 degree hitch angle (I will check that this evening) I estimate the horizontal component Fh as Fr/cos(theta) where theta is the hitch angle. With that the Fh becomes 406 lbf for the stone boat, 549 for the tongs and 565 for the choker chain. Then, based on the load weights the observed friction coefficient is 0.33 for the boat, 0.52 for the tongs and 0.54 for the choker chain. Then, and I will take a leap here to keep from killing anyone other than you who reads this, to find the optimal hitch angle one has to find the angle theta which provides a minimum pull. This occurs when the first derivative of the pulling force, P, with respect to theta, equals 0. This occurs when the value of the kinetic friction coefficient = tan(theta).

So an optimal hitch angle for the stone boat is 18 deg., 27 deg. for the tongs and 28 deg for the choker chain. Those are close to your estimates and you are right that an optimal hitch angle is different for different hitching methods. And you are correct that in most cases we cannot hitch at an optimal angle because of physical constraints so the recommendation is to hitch as close as possible. I think it is cool that an optimal hitch angle for a wagon for instance with low motion resistance is quite low because it is low anyway with the high hitch point. An optimal hitch angle for tongs with greater resistance is greater which works out nice because the hitch point is lower. It is nice when the theory and practice are consistent, gives you reason to think that there are opportunities for real and acceptable changes with good ideas.

Rock on, draft animal nerd. 😀

[Andy Carson]

Hey, who’s calling me a nerd! Need I point out the you just performed calculus on a trigometric function, and in public??? 😀

Thanks for the thoughts, numbers, and math, Tim. It is interesting that we come up with functionally similar angles using different methods. Getting similar answers with different methods definitely gives more confidence. I am probably picking a nit, here, but I like an analysis that treats the front of the log differently than the back of the log because the front can dive into the ground and till while the rear cannot. Because of this fundamental and important difference, I think the friction coefficient for the front of the log will be different from the rear. Also, I don’t think that a lift to one side of a log will necessarily shift that same weight to the rear. A lift of 525 lbs on one end, for example, doesn’t make the other end weigh 1050 lbs… So, I’ll rework this with the additional information. I kinda wonder why, because I’ll probably end up with the same take home information as in our previous two posts, but let’s see…

Sled: As the front of the sled can’t till and the rear of the log is off the ground anyway I’ll estimate draft without reguard to front and rear ends. For a pull of 392 lbf at 15 degrees, the horizontal component is 379 lbf (cos15degrees*392), and the vertical component is 101 lbf (sin15degrees*392). The vertical component removes 101 lbs from the weight of 1225, yielding an applied weight of 1124. This gives a friction coefficient of 0.34. with this friction coefficient, one can plot the resulting forces (shown in the graph below). I get a minimum draft force at 19 degrees, which is damn close to Tim’s 18 degrees, but what really jumps out at me in this plot is just how flat it is. All resulting draft forces from an angle of 7 degrees to 29 degrees are within 2% of the minimum. So, for all practical purposes, the angle of draft doesn’t matter for a sled.

Tongs: I am going to model the rear of the sled as having a friction coefficient of 0.4 (slightly more than the sled) because it can’t till and has a rough surface (and because I’m a stubborn bastard). The downward force of the rear half is still 525 lb, resulting in horizontal drag is 210 lbf. For a pull of 549 lbf at 15 degrees, the horizontal component is 511 lbf (cos15degrees*549), and the vertical component is 137 lbf (sin15degrees*549). The vertical component removes 137 lbs from the front weight on 525, yielding an applied front weight of 388. As the horizontal component of the pull is 511 lbf, and the rear takes up 210 lb of this, the resulting friction coefficient is 0.78. Wow, can the front dig in! With the friction coefficients of the front and rear, one can plot the resulting forces (shown in the graph below). I get a minimum draft force at 38 degrees, which is a ways away from Tim’s number due to substantial lift on the front (not the rear) of the log. This probably doesn’t matter is practice, though, because either angle is challenging to achieve. Here the graph illustrates how much more sensitive this setup is to draft angle, with a 2% increase over minimum draft at 27 degrees, a 5% increase over minimum draft at 20 degrees, and 10% increase at 13 degrees.

Chain: similar to the tongs, I am going to model the rear of the sled as having a friction coefficient of 0.4. The downward force of the rear half is still 525 lb, resulting in horizontal drag is 210 lbf. For a pull of 565 lbf at 15 degrees, the horizontal component is 546 lbf (cos15degrees*565), and the vertical component is 146 lbf (sin15degrees*565). The vertical vector yields an applied front weight of 379. As the horizontal component of the pull is 546 lbf, and the rear takes up 210 lb of this, the resulting friction coefficient is 0.89 for the front end of the log (which is similar to the friction of a car tire on pavement). Wow again! Just as in the other examples, I can use the friction coefficients to plot total forces at different draft angles. For the chain, I get a minimum draft force at an impractical 42 degrees. This type of hitch is even more sensitive to draft angle, with a 2% increase over minimum draft at 30 degrees, a 5% increase over minimum draft at 23 degrees, and 10% increase at 16 degrees.

[Tim Harrigan]

@Countymouse 28145 wrote:

“…I like an analysis that treats the front of the log differently than the back of the log because the front can dive into the ground and till while the rear cannot. Because of this fundamental and important difference, I think the friction coefficient for the front of the log will be different from the rear. Also, I don’t think that a lift to one side of a log will necessarily shift that same weight to the rear…”

I agree that the friction from the front and back are different, I just have trouble picking the functioning system apart in that way. It is interesting to do an ‘what-if’ analysis but I prefer to start from the measured values and go from there. When you re-integrate your front and rear-of-log solutions do the actual measurements validate your approach with the tongs or choker? I agree that the weight transfer itself is not significant, after all the rear end of the log will not carry 1050 lbs until it is standing on end. But small shifts are likely to cause considerable increases in drag on the end of the log.

“….Sled: …what really jumps out at me in this plot is just how flat it is. All resulting draft forces from an angle of 7 degrees to 29 degrees are within 2% of the minimum. So, for all practical purposes, the angle of draft doesn’t matter for a sled…”

Yes, I agree, so much so that one would be justified to ask ‘Who cares?” It always comes back to point that we want to harness the animals in a way that allows them to apply their power comfortably and efficiently. There is a lot of room available to do that.

“…Tongs: the resulting (front) friction coefficient is 0.78. Wow, can the front dig in!…”

Yes, possibly, but you will notice that I angled off the front of the log to prevent that from happening. So I doubt that this is realistic in this case. Again, do the measured values verify the results for your reintegration of the front/rear components?

“..The graph illustrates how much more sensitive this setup is to draft angle, with a 2% increase over minimum draft at 27 degrees, a 5% increase over minimum draft at 20 degrees…”

“…Chain: …the resulting friction coefficient is 0.89 for the front end of the log (which is similar to the friction of a car tire on pavement). Wow again!…”

These coefficients seem quite high and would indicate considerable ground disturbance or plowing of snow if in the winter.

Again, I think the physics of hitching is interesting and instructive, but the bottom line is it has to allow the animal to apply their body mass and strength in a way that is comfortable and efficient. We know that it means a low hitch point for heavy resistance. Also, because of variations in terrain, etc., I think an actual optimal hitch angle would probably be a few degrees greater than a calculated optimum. So observation and teamster skill in adjusting to the situation are key components.

[Tim Harrigan]

Andy, check out post #26 in the draft buffer thread of the working with draft animals section. Look at the graph comparing pulling forces for a sled at various hitch angles and with the center of the load shifted from front to center to back. I think you will find it interesting.

[Andy Carson]

Yes, post #26 is very interesting… I will have to give this a good think and see if this data contradicts or supports my modeling… I agree that without “real” measurements, we are kinda arguing about how many angels can dance on the head of a pin. I think this kind of data analysis is sometimes (maybe even often) useful though, because it can identify which parameters are most interesting to test in the field. Sometimes, too, the mathmatical modelling can suggest interesting and potentially useful modifications. I am sure I will have more thoughts when I think through and compare post #26.

[Andy Carson]

Draft buffer post #26 has made me think a lot. One of the things that this brings to mind is that we have been modeling the soil as an incompressible substance, even though we know better. Now when I say compression here, I don’t mean irreversible compression, just the “give” that you feel when you step on soil versus say concrete. I think this is an important factor, and here’s why. I think Tim is right when he points out that my friction coefficients for the front of the log are too high to be reasonable. And yet experience, observation, and Tim’s tractor+sled data (post 26) would indicate that weight in the front somehow causes more drag than weight in the back. If one doubts this, one can also look at the wear on sled runners. I have had to replace mine several times, and the most wear is invariably at the leading edge of the runner right where soil contact is made. There’s a good bit of wear at the very rear of the runner too, but the middle of the runners always holds up a lot better. Now every part of those runners sees exactly the same terrain, travels the same distance, and is made of identical materials, but the wear pattern is dramatically different. On my sled, I bet the leading edge wears at twice the rate of any other area. Maybe even more than twice… So, because the friction coefficient is probably almost the same (soil vs steel), there must be more weight on the leading edge somehow. With even or variable loads, that can’t make sense without some degree of soil sinkage. With soil sinkage, the front of the sled would be thrust up onto upcompressed ground and be briefly suspended between the leading edge of a flat runner and the trailing edge. At this point, weight from the front half of the sled will compress the soil along the leading edge, and the trailing edge will sit on already compressed soil at the rear of the sled. The elevation of the sled’s front may or may not be noticeable (just like you can’t always see the front of an icebreaker elevate) but the concept of “lift and fall” still applies.

So, I’m going to see if vertical lift and the work of compressing soil can account for the increased drag from the front of the sled in this test (and see if this also works for the sled in draft buffer post #26). During the “lift” component a pull into compressible soil with a flat bottom sled, all weight will be suspended between the leading sled runner edge (or other contact point) and the trailing rear edge. The downward force on the leading edge has to overcome the compressibility of the soil at a rate proportional to the angle of the leading edge. Half the load in Tim’s example weighs 613 lbs (1225/2), which is elevated by the the 15 degree total pull of 392 lbf. The vertical component of this pull is 101 lbf, leaving 511 lb on a 10 degree incline (estimated angle of attack) and a horizontal pull of 379 lbf. Moving 511 lb up a 10 degree incline takes 89 lbf along at a 10 degree angle. That equates to a 90 lbf horizontal pull. So 90 lbf out of 379lbf might used up compressing soil along a 10 degree leading edge of a sled. That leaves 289 lbs remaining. There is still 511 lbs in the front and 613 lbs in the rear, which if applied equally would yield friction coefficients of 0.26. Now these friction coefficients are really quite low, but this is analysis models this interaction as non-compressing, non-tilling, and smooth, and takes the leading edge into account separately, so maybe this isn’t stange… Overall, using this analysis the front edge is responsible for about 60% of the draft and the rear is responsible for about 40%. This general pattern explains some of the patterns seen in post #26, where loading up the front of the sled resulted in higher draft forces than loading the rear. It’s really pretty tough to do these analysis in great detail, because so much depends on the exact angle of attack at the leading edge, the length of the sled, and soil properties (such as compressibilty and dampening). Because of these complications, I can’t compared exact numbers, I can really only see if the trends match, which they do.

There are practical applications of this modelling. One aspect that comes out to be important is how critical a shallow angle of attack is on sled runners. Also, it would be advantageous to extend the runners out in front of the load so they are not as weighted while they are being “lifted.” Lastly, having a very shallow rounded bottom to the runners (very slightly higher in the middle) will prevent much of the the deleterious loading of the front end. I bet a runner such as the top runner in the photo below would be much more efficient than the runner below. Maybe as much as 10% improved with respect to draft for a given weight… Now, there’s something to test! 🙂

[dominiquer60]

This is the nerdiest thread yet, I can’t follow the details, but I like it:)

[Andy Carson]

Sorry Erika,
sometimes I get excited and do a bunch of modeling, thinking, and writing and don’t stop to explain why I’m doing it in the first place…

Basically, I think that Tim’s data in this post and others can be partially explained by having greater drag from the front of the sled or log than from the back. I think this makes some intuitive sense and matches with experience and observations but it is difficult to explain exactly why that is. I am interested in trying to figure out “why” because whatever is causing the additional drag might be able to be “fixed.”

Most of the math is me coming up with theories of what might be causing the additional drag in the front, and seeing if this math matches with measured values. The idea is that if the math predicts values that are far from the measured values, than the model/theory is probably wrong. I created a model based on different coefficients of friction that explained the values in this thread alone. Tim pointed out (very gently I might add) that although it does match the value in this post, the coefficients of friction are hardly believable and it the model doesn’t match measurements in post #26 of the draft buffer thread. So, I came up with a different way to explain exactly how the drag might be higher in the front than in the rear, based on the compressibility of the soil. I think it makes sense, and explains the data in this post and most of post 26, but we’ll see what people have to say… It sounds like whatever the explaination is, it is probably going to be somewhat theoretical. That’s what’s nice about the sled runner shape test. If a theory can’t be tested somehow, there’s no way to know if it’s true. Also, the runner design is a practical application of the concept, if it does indeed work. There will probably be some batting back and forth of ideas to come, that is if anyone else wants to be nerdy.

[Tim Harrigan]

@dominiquer60 28153 wrote:

This is the nerdiest thread yet…:)

You can blame that on Andy. :rolleyes:

[Tim Harrigan]

Andy, I have come to the same conclusion, the front of the runners compress and firm the soil as the implement moves across it so the back of the sled is running on firmer and smoother soil and thus a somewhat lower friction coefficient. It is like a wagon on soft ground, the front tires firm the ground and reduce the rolling resistance for the rear tires running in the wheel track. The simplified approach to predicting draft based on a load and the surface has the advantage of being straight forward but it lacks precision if you really want to understand the details. In fact, the approach does not account for the surface area of the sled or boat so in theory you could take the load off the boat, turn it on edge and reload and the draft would be the same. That is the incompressible assumption that we know does not hold, but it is pretty close in some cases.

I compared a stoneboat with a sled with 4 inch runners traveling across firm hay ground, soybean stubble in the spring and then a recently planted oat field. The sled and boat were identical on the firm hay ground, sled draft was somewhat higher than the boat draft on the soy stubble, and considerably higher on the oat ground. So that was a clear demonstration of the compression and compaction of the soil where the reduced flotation from the reduced runner surface area increased sinkage and increased draft.

I have thought about changing the surface material but not the shape of the runners. You rockered the runners, but you also seemed to describe a natural process of wear whereby the runners took on a rockered profile. So do the runners naturally develop a low-draft profile? Are you suggesting we need to be more attentive to mimic the shape of worn runners when we replace them with new runners? That makes sense to me.

I have wanted to test an implement pulling a load over the same surface but with increasingly greater loads. I suspect that under the same conditions, a pulling contest for instance, when the sled is loaded to a % of the team weight, the bigger team pulls a greater portion of their weight e.g. experiences a higher friction coefficient than the smaller team. Because of the compressibility of the soil and the work needed to compress the soil. It is not something you can see, but I suspect it could be measured.

[Author]

Posts