Testing the Trace Harness
Pro or Contra

by Paul Schmit of Luxembourg

In today’s times, people can exchange information within seconds worldwide. Nevertheless, all the modern digital communication channels cannot prevent something valuable from being lost, the knowledge and skills from earlier times, if the practicing in real life situations is missing. This also counts for the innumerable work horse harness variants that were originally adapted to the prevailing working conditions in the different places around the globe.

However, sometimes the various techniques from bygone times experience a renaissance and are brought back to life and partly adapted to the modern level of development, mostly by trial and error. What is characteristic of this is that the re-invention is often presented as superior to others and groups of supporters and opponents then form all too quickly.

An example of such a hotly debated item in Central Europe is the so-called “trace harness”, also referred as “long gears”. In the bygone time, this type of harness was not only used for agricultural or forestry work, when extra power was required to pull a heavy load, but also in vineyards and in vegetable production. In the first-mentioned operation, two or more horses were hitched “in-line”, one behind the other, and the fact that there was no singletree dragging on the ground prevented the extra lead horse(s) from stepping over the traces. In row crops, the reasons for use were to avoid damage to the crops, by a low singletree, and to be able to turn more easily and very sharply at the headlands.

This type of harness was also widely used across Europe in railway operations. In marshaling yards, the so-called shunting horses were used to separate and assemble wagons into complete freight trains for different routes. In the US, this activity is known as switching. Here too, the constant tight turnings and need to overcome obstacles such as railway tracks or switches, without getting caught by the singletree, were the reasons for choosing this special type of harness.

Mid of the last century, in most European countries, the decline of work horses for railway operations, as with the farming, was driven by the replacement through motorised equipment. This was not a process that happened immediately. The Second World War, and the rationing of petrol meant a temporary halt to the reduction in horse numbers, but it was only temporary. After the war ended, the push to mechanise gathered pace and by the end of the 1950’s there were very few railway horses, if any still working.

Unlike for cartage, work horses used for shunting remained in use a while longer. But this era, that had lasted over 100 years, came also to its end. In Great Britain, probably the last of them all, a Clydesdale gelding named Charlie, retired on February 21, 1967, from Newmarket Railway Station. Charlie achieved some celebrity as the last working railway horse, when British Pathé, a producer of newsreels and documentaries, filmed him at work in 1964. This short film can be found under the following link: https://www.youtube.com/watch?v=kQ9DrerydpM

The first two pictures were taken at main railway stations in Sweden. Alvesta as the junction between the Southern Main Line connecting Malmö to Katrineholm and the Coast-to-Coast Line connecting Gothenburg to Karlskrona. The Malmö railway station was the southern terminus of the Southern Main Line at that time. What makes these pictures special are the breast collar harnesses used, in contrast to forestry and agriculture, where full collars were in use. Since the hitch point on the railway wagons was relatively high, it can be assumed that this hitching method was regarded as a more efficient draft power transmission.

Following British harness traditions, both chain traces are kept apart by a so-called spreader bar, fixed behind the horse’s hind legs, for joining further back to the hook hitched to the load. This differs from the harness on the Continent, where leather or rope traces are fixed to a high-mounted singletree to be connected by a single chain to the load. Both, the spreader bar as well as the singletree are kept in position by so-called hip-straps in such a way that chafing of the hind legs or interference with the walking action are minimized, at least that’s the theory.

And this is exactly the point where the story of supporters and opponents, described at the beginning of this article, comes full circle. Proponents praise the efficiency of this harness type, as it eliminates the need to constantly raise the singletree during empty runs and turning as well as reversing is as simple as it can be. This is especially true when used for railway operations, but also in the forest, where dragging brushwood or getting caught behind tree stumps is unavoidable if the singletree drags on the ground. Opponents do not question the human’s comfort and efficiency at work but argue that the additional load on the horse’s croup as well as the constant contact between hind legs and singletree, when moving without load, can lead to health problems for the animal.

Reinvent the tried and tested

The late Matthias Rensing, one of the most recognized full-time horse loggers in Germany, deserves the credit for reviving this type of harness, at least in Europe’s German-speaking countries. Building on the practical experiences of loggers from France, he began to develop in 2010 his own concept together with the German harness manufacturer Sattlerei und Pferdebearf Maus, which is nowadays sold under the name of “Schwebeortscheit UM 4 – System Matthias”.

The German name “Schwebeortscheit” translates as “floating singletree” and indeed, contrary to the harness of the bygone time, where the adjustable hip straps were made of leather, this new variant utilizes rubber bands, which allow some play in the suspension of the singletree. The upper leather straps, carrying the whole assembly, cross over the horse’s croup and are adjusted so that the singletree rests over the hock, exactly on the back of the horse’s knee, when the horse is walking without load. Additionally, the original French harness design uses a wide leather cushion, below these straps on the croup, to better attenuate possible pressure. Usually, the tail is tied up and bandaged to prevent it from getting caught in the singletree.

In France, this harness type is named “palonnier porté” or “bas-cul”, which refers to the fact that the singletree is carried below the horse’s hind end. The former French manufacturer of railway equipment Decauville, founded in 1875, sold it under the name “l’écrevisse” and their design was already very close to the modern German version described here-above. Following a manufacturer’s catalogue from 1904, this harness type was not only in use for railway operations, logging or hoeing, but also widespread in quarries or mines as well as for horseboating.

A further development in the modern German harness are the traces. Matthias Rensing was always an advocate of draft dampers and had synthetic rubber dampers recreated, originally developed by the Swiss Claude Fanac in the 90s and sold at that time under name ZEP – Zugerleichterung fu?r Pferde (draft relief for horses). Here, a rubber core is braided into a fiber rope, made of polypropylene, and the spring/damper effect results from the contraction of the rope under the load and the friction between rope and core. Compared to steel draft springs coupled to the traces or incorporated in the singletree, this solution is characterized by a very low weight and no additional space requirement. Here again, many people praise, or query, the effect of these parts in the harness.

The difference between presumption and knowledge is …

… information, and from 2011 to 2018 the NGO Schaff mat Päerd carried out various trials with different traces and spring/dampers to get to the bottom of this matter. The dynamic tests took place under real working conditions during logging and hauling firewood with a Swedish eight-wheeler logging wagon. Furthermore, static tests were done to establish force-elongation diagrams for each trace and trace-spring combination. Finally, without being exhaustive, it was concluded that a lot of the statements about the benefits of draft springs and dampers are more wishful thoughts than scientifically reliable findings.

Building on this database, the following hypotheses were formulated for new trials with trace harness:

Hypothesis 1
In addition to the static load, for carrying the singletree, its rubber band suspension creates some additional pressure on the horse’s croup, during the pulling effort, depending on the draft effort.

Hypothesis 2
The singletree, mounted in the middle between the backband and the log, increases the draft force oscillations during the pulling effort, caused by its own up and downswing, compared to a singletree mounted closely to the log.

Hypothesis 3
The damping capacity of the draft force oscillations by the rubber damper in combination to the synthetic traces is limited and cannot prevent draft force peaks during logging.

To verify or refute these hypotheses, two test series were carried out on May 10 and June 6, 2024 in cooperation with the Horsepower Competence Center, run by the Robbesscheier Public Nature Center, located in the North of Luxembourg, which makes part of the hilly and richly forested Ardennes, the original breeding area of the Ardenner horse.

For each of the trials, the same horse named Poli de Racelle, a 10 year-old Ardenner gelding with Belgian origins and belonging to the Horsepower Competence Center, was used. The measuring device consisted of the following components:

• two Lorenz K-12 force sensors mounted between the singletree and both of the synthetic traces
• one Lorenz K-12 force sensor mounted between the singletree center and the hook
• two Lorenz K-100 force sensors mounted between the singletree and both of the rubber support straps
• one AHLBORN Almemo 2690-8 data logger with memory connector and micro SD card mounted on leftside hame

The measured values were recorded with a summary measuring rate of 100 Hz resulting in a frequency of 17 Hz for each of the six dimensions (time, left-side, right-side and total draft forces as well as left-side and right-side support forces).

In addition to the total draft force, the following graphs only show the left side in order not to overload the graphs with color. The preliminary tests, mentioned above, showed already that the force relationships are almost symmetrical at the singletree, and this was confirmed again by the measurement setup, discussed here, with five force sensors on the singletree.

True or False

For Hypothesis 1, with the load as the main influencing factor, three different logs were pulled on a flat and short-cut grassland belonging to the Robbesscheier Public Nature Center. Graph 1 for a light log of just 36 kg could lead to the false conclusion that the rubber bands transfer a part of the draft force because they are more heavily loaded with 0,080 kN than the traces with only 0,048 kN per side on average. But a comparison with the graphs 2 and 3 for two light logs with a total weight of 69 kg and a heavy log with 188 kg weight, shows that the load on the rubber band suspension of the singletree remains almost constant with on average 0,087 kN and 0,085 kN. Furthermore, as all three graphs show, the support load in the rubber bands is very dampened compared to the load on the traces.

The static load for carrying the singletree without the horse moving forward was determined to 0,042 kN per side, corresponding to a total load on the horse’s croup of 8,6 kg. The weight of the singletree suspension must also be considered, which puts an additional 1.9 kg on the horse’s croup. The own weight of the singletree with hook is only 5,35 kg. The fact that the static load on the croup is higher than the own weight of the singletree can be explained by the angle of the rubber band suspension to the rear. If the rubber band were to run downwards in a vertical line, the load on the croup would correspond to the sum of the weights of the singletree and its suspension.

In conclusion for Hypothesis 1, it can be stated that the assumption that the load on the horse’s croup is variable, and depending on the load to be pulled, was refuted. However, this load is twice as high on average as when the horse is at rest. No veterinarian has been able to answer whether the additional total load of around 19 kg on the horse’s croup harms the horse, as there is a lack of scientific knowledge on this matter. This load corresponds exactly to the weight of the traditional collar with wooden hames, which the test horse was equipped. Traditionally, the collars in the Benelux countries and the German Rhineland were rather heavy compared to other European harness. A complete Scandinavian harness, including collar, hames, back pad and complete breeching, only weights the same amount. However, the contact surface between the collar and the horse’s neck is bigger than between the hip straps and the horse’s croup. Furthermore, the horse’s anatomy is different at both places.

For Hypothesis 2, the synthetic traces were extended by chains so that the singletree was in the usual position close to the log. For this trial, the rubber band suspension of the singletree was replaced by leather trace carriers. The length of the extension chains was chosen so that the draft angle was the same as in the previous measurements. Only the heavy log was pulled on the flat and short-cut grassland.

By comparing graphs 3 and 4, it can be concluded that draft force oscillations are 22% lower when the singletree is placed close to the log, instead as floating singletree close to the horse. By that, Hypothesis 2 is partially confirmed, but here too there are no comparative values to quantify the stress on the horse.

For Hypothesis 3, five different logs, which could be rated to medium and heavy weight, were pulled in a downhill forest belonging to the Robbesscheier Public Nature Center, from the felling site, between still standing trees, to a forest track.

The required draft effort was significantly higher here, as expected, compared to the flat and short-cut grassland, and corresponded to the usual working conditions of logging horses in the hilly Ardennes forests.

The support forces for the singletree remained on the same level or even lower, compared to the trials on the grassland. For a heavy log, a difference of 14 % could be measured. The slightly lower value can be explained by downhill working. Also, no excessive oscillations in the support forces could be measured.

As the preliminary static tests, described and illustrated above, showed, synthetic rope traces with rubber dampers present some considerable elongation under the load. Unlike steel draft springs with a linear characteristic, the synthetic trace and damper assembly is characterized by a much smoother progressive curve, as shown in graphs 5.

For the maximum force of 0,8 kN and elongation of 45 mm for a steel draft spring in each trace, which corresponds to a total draft force of 1,6 kN on the singletree center, the synthetic trace/damper combination shows with 86 mm nearly the double in elongation. Following graphs 6 and 7, draft force peaks of 2,2 kN, corresponding to the maximum load of the synthetic trace/ damper assembly, occur frequently. Here, the assembly has expanded by 120 mm.

This lengthening not only changes the initial setting of the trace length, chosen in static state to have a straight draft line between the collar and the singletree, but also has an influence on the draft force oscillations. With reference to the graphs 6 and 7, it must be concluded for hypothesis 3 that the synthetic trace/damper combination did not adequately dampen the draft force peaks. For a medium heavy log, peaks per side of 3,530 kN could be measured, corresponding to a total load on the horse’s shoulders of 720 kg. For a heavy log, these values reach even 5,498 kN or 1120 kg impact on the horse’s shoulders.

Many thanks go to Nora Schaul and Charel Braquet of the Horsepower Competence Center, who made this study possible.

Paul Schmit
Schaff mat Päerd asbl