An Optimal Braking Force Distribution in the Rigid Drawbar Trailers with Tandem Suspension
Online veröffentlicht: 31. März 2025
Seitenbereich: 94 - 105
Eingereicht: 15. Nov. 2023
Akzeptiert: 15. Nov. 2024
DOI: https://doi.org/10.2478/ama-2025-0012
Schlüsselwörter
© 2025 Zbigniew Kamiński, published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
The design of the running gear of agricultural trailers with rigid (unbalanced) drawbar depends on their load capacity. Semi-mounted trailers with laden weight of up to 13 tonnes have a single axle, up to 19 tonnes tandem axles and up to 22.5 tonnes triaxle [1]. A multiple-axle unit consists of axles spaced closely together, usually between 1.2 m and 1.85 m [2, 3]. Tandem and triple axles are used to increase the vehicle's load capacity and load distribution between the axles, regardless of road surface irregularities [4]. Generally, two basic tandem suspension arrangements have been developed: a central pivoting single vertical semi-elliptic or parabolic spring which has an axle clamped to it at either end and a pivoting reactive or non-reactive balance beam which interconnects adjacent first and second semi-elliptic springs via their shackle plates [5,6,7].
During a brake application, all the vehicles of a combination tractor-trailer should be braked with similar intensity to enable efficient deceleration without the risk of the combination losing its directional stability. From 2016, EU legislation on agricultural vehicles [8] has required agricultural trailers travelling at speeds above 30 km/h to comply with the same braking efficiency of 50% as commercial trailers [9]. Furthermore, agricultural balanced trailers with a total mass exceeding 3500 kg (categories R3 and R4) require a specific brake force distribution among the axles. As in the case of commercial vehicles [9], the individual parts of the combination are treated as individual vehicles, so that the coupling interactions between them are not taken into account. No recommendations are made in this respect for semitrailers. To share the brake force distribution between a tractor and a towed vehicle, permissible compatibility corridors for the braking rate of the tractor and the towed vehicle refer to the pressure values of the control line between these vehicles that have been introduced. The compliance with compatibility requirements, as well as requirements regarding high-speed operation (response time of less than 0.6 s [8]), contribute to the shortening of stopping distance of tractor-trailer combinations and the reduction of forces in the coupling during emergency braking [10]. The implementation of the new European legislation in the field of agricultural vehicles places high demands on the manufacturers of agricultural trailers, tractors and machines concerning braking systems [11].
Most of the works on agricultural trailer braking deal with various aspects of the braking process of tractor-trailer combinations, mainly with two-axle trailers [12,13,14]. Papers [15, 16], describe the braking mechanics of a tractor with a single-axle trailer while moving up and down a slope under various operating conditions. The dynamic behaviour during braking of tractor-semitrailer combinations in terms of stability and road safety has been analysed in [17] and [18]. Paper [19], on the other hand, describes the braking of a semitrailer endowed with an inertial braking system, working in aggregate with the tractor. Both the theoretical and practical aspects of the braking performance of tractor-semitrailer combinations were examined in [20] and [21]. To simplify the theoretical considerations in [20], the tandem bogie suspension was replaced by a single axle and the forces from the wheels were applied to the bogie joint with-out considering the interaction between the tandem axles. From the designer's (manufacturer's) point of view, any analysis of the braking mechanics of a tractor-trailer combination seems to be of little use for the design of semi-trailer braking systems, as it requires specific assumptions to be made about the geometric and mass parameters of a farm tractor. Moreover, these considerations usually apply to single-axle semi-trailers but not to semi-trailers with tandem suspension. And as is well known [3, 4, 22], when a braking force is applied on a tandem axle, there is often a load transfer among the axles, and the lighter loaded axle tends to lock up before the other. This phenomenon, which is dependent on the tandem suspension type used, harms braking performance and directional stability. If the lock-up occurs on the trailing axle, it can lead to a complete loss of directional stability [3]. Therefore, in this study, the influence of inter-axle load transfer in semitrailers with different tandem suspensions has been considered in the development of the brake force distribution methodology for use by designers and manufacturers of agricultural machinery, thus filling a gap in the literature.
The design process of a new brake system starts with the selection of the distribution of brake forces [23]. The axle load transfer and braking force distribution play an important role in the safety and dynamic stability of road vehicles [24, 25]. In general, for the correct design of a vehicle braking system, it is essential to know the ideal brake force distribution between the axles for laden and unladen vehicles [26]. With the braking distribution at the ideal level, the braking performance is maximized, and the brake stability is then also guaranteed [27]. The optimum braking condition is achieved when each axle has the same utilised adhesion, i.e. the ratio between the braking force on each axle and its vertical load is the same [28, 29]. Under such conditions, the ratio of the longitudinal force and vertical force on the coupling device is also the same [30].
This latter condition became an inspiration to use the brake force distribution strategy for two-axle trailers, according to the EU regulations [8], to determine the braking force distribution in semi-trailers between the axle unit (single or tandem) and the coupling device that attaches the towed vehicle to the farm tractor. With this approach, a semi-trailer can be considered as if it were an individual braked vehicle with braking forces at the axle wheels and the coupling device.
To calculate the brake distribution of a single-axle semi-trailer, the analytical method can be used as for two-axle trailers [31, 32]. However, for tandem semitrailers, the analysis is more complicated because the leading and trailing axle loads are functions of the trailer load and the type and geometry of the tandem axle unit [4]. Therefore, even for vehicles with the simplest of tandem suspensions, such as a walking beam and bogie [31, 33] or a double elliptic leaf spring suspension [34], optimisation methods are used to calculate and select the brake force distribution. To find an optimal linear force distribution, which is mostly used in trailer air brake systems due to the linear characteristics of the brake force regulators, the quasi-Monte Carlo method was used in this paper.
The remainder of the paper is structured as follows: in section 2, the equations of forces acting during braking on a tractor single-axle semi-trailer and a tractor tandem-axle semi-trailer combination are developed to find the condition of ideal braking force distribution; in section 3, the analysis of the forces acting during braking on a semi-trailer with different types of tandem suspension is presented; in section 4, the UE regulations for the brake force distribution adapted to semi-trailers are described; in section 5, the quasi Monte Carlo method and the algorithm of linear brake force distribution are described. The results of optimising the different tandem axles are analysed and discussed in section 6. Finally, a summary and conclusions are drawn in Section 7.
The findings of the work are expected to provide a reference for the design and evaluation of the air braking system of agricultural semi-trailers, especially with different types of tandem axles, to improve braking performance and reduce coupling forces.
Tractor-semitrailer combination is modelled as two rigid body hinges to each other, and suspension deflection is ignored. The forces acting on a braking farm tractor with single and tandem axle semi-trailers are shown in Figures 1–3. For simplicity, it is assumed that aerodynamic and rolling drag is neglected.

Forces acting on a farm tractor (ISO coordinate system [35])

Forces acting on a single-axle semi-trailer

Forces acting on a tandem-axle semi-trailer
Using the notation from Fig. 1, the equations of forces and moments exerted on the decelerating tractor are of the form:
The equilibrium equations of forces and moments exerted on the single-axle semi-trailer shown in Fig. 2 are as follows:
From the equations (2) and (5) of vertical forces (after elimination of the reaction R1) and the equation of moments for the tractor (3) and the trailer (6) concerning the coupling, the vertical reactions are obtained:
In the case of ideal braking of a tractor with a single-axle semi-trailer, the adhesion utilisation of each axle is the same and equal to the braking rate of the combination, i.e.
So that during ideal braking, the ratio of the tangential force to the vertical force acting on the coupling:
The system of equilibrium equations for a braked single-axle semi-trailer with walking beam tandem suspension (Fig. 3) is given as:
From the equations (2), (17) of vertical forces (after elimination of the reaction R1) and the equation of moments for the tractor (3) and the trailer (18) with respect to the coupling and the relationship between the reactions in the tandem suspension:
During ideal braking of the combination of tractor-semitrailer with tandem suspension, the adhesion utilisation of each axle is the same and equal to the braking rate:
The quotient of the tangential force to the vertical force on the coupling:
Thus, the ideal braking condition is unambiguously defined and to determine the ground reaction and then the distribution of braking forces, it is not necessary to analyse the braking process of the entire combination - only the braking process of the trailer alone may be considered.
A rigid two-dimensional model is used to analyse the brake force distribution of tandem-axle semitrailers (Fig. 3). The braking forces T
The simplest form of tandem unit is a walking (rigid) beam [37] mounted pivotally to a frame hanger on either side of the vehicle (Fig. 4-a). In the bogie suspension [38], parabolic tapered springs are anchored upside down to the trailer frame in place of walking beams (Fig. 4-b).

Forces acting on a walking beam (a) and bogie suspension (b)
The forces exerted on the walking beam and bogie suspension during braking are shown in Fig. 4.
The same equilibrium force and moment equations can be used to describe both tandem suspensions:
By solving the system of equations (17), (18) and (31) together, considering from equation (16) that
In tandem leaf spring suspension, the two most common spring types are the double eye leaf spring and the slipper spring. For agricultural trailers, the second type is more common [37,38,39]. The front eye of both the leading and trailing springs are hinged directly to the front hanger bracket and the levelling beam, respectively, through pin joints (Fig. 5). The rear end of the springs is captured in the equalizer beam or rear hanger.

Forces acting on a two leaf spring suspension
The forces applied to the two leaf spring suspensions with two unsprung weights are illustrated in Fig. 5.
For the unsprung weight
The equations of the balance of forces and moments for the unsprung weight of the
Equations (37) and (40) give the reactions at the equalizer beam ends:
After substitution of the expressions (41) and (42) into the equilibrium equation of the force moments on the equalizer beam:
Another version of the tandem axle configuration uses only two springs with slipper ends. Vertical forces are transmitted to the trailer frame via the front and rear hanger brackets and equalizer beam (according to the BPW equalising beam) [37,38,39].
Longitudinal forces are transferred by connecting the radius rods between the axles and the front and middle hanger brackets (Fig. 6). The parameters

Forces acting on a two leaf-two rod suspension
The force and moment equations for unsprung weights
From equations (49) and (52), the reactions acting on the ends of the equalizer beam are given:
Solving equations (17), (18) and (53) – (54) together, the semi-trailer axle loads, and the coupling force are obtained:
Equations (56) – (58) become much simpler when
A tandem double leaf spring suspension with equalization [4] has a pair of slipper springs and a mechanical equalisation of the braking load (Fig. 7). The rear end of the front spring is connected to the rear end of the rear spring by a rocker arm which is hinged to a central hanger bracket. This rocker distributes static (and shock) loads evenly between the two axles. An alternative design solution for a non-reactive tandem suspension with a bell crank and a tie-rod linkage is described in [5]. The forces acting on the two leaf spring suspensions with equalisation are presented in Fig. 7.

Forces acting on two leaf spring suspension with equalization
The force and moment equations for the unsprung weights
Equations (61) and (64) are used to compute the reactions at the ends of the beam:
By solving the system of equations (17), (18), (65) ‒ (67), the vertical coupling force and the dynamic axle loads are obtained during deceleration of the semi-trailer:
With air suspension, the air springs are mounted to the trailing arms via a crossmember and attached to the frame at the top (Fig. 8). The trailing arms are hinged to the hanger brackets and axle housings. All the air bags are connected by air lines to balance the axle loads. Vertical forces are distributed between the hanger brackets and the airbags. Longitudinal forces from braking are transferred to the trailer frame through the hanger brackets.

Forces acting on air tandem suspension
The balance equations for the forces and moments applied to the suspension with unsprung weights
On the assumption that the pressure in the airbags is the same, the vertical reactions transmitted by the air springs will be the same as well:
The solution of the system of equations (17), (18) and (77) is the vertical coupling force and the trailer axle loads:
As proved in section 2, the ideal braking condition for semi-trailers is achieved when the rate of utilized adhesion of each axle and the quotient of the horizontal to vertical force at the coupling is equal to the braking rate z of the combination. For semi-trailers with a single or tandem axle, this condition can be expressed as follows:
The adhesion utilization rates used by the front coupling and the rear axle assembly are calculated based on the relationship:
With ideal braking, stopping distance is minimised and braking efficiency requirements are met with reserve (
Due to varying trailer loads, it is virtually impossible to ensure ideal brake distribution, even with the application of brake force regulators. Therefore, for agricultural trailers travelling at speeds above 40 km/h, acceptable limits have been set for the deviation of the adhesion utilisation rates of individual axles from the optimal distribution. When considering brake force distribution, each part of the combination is deemed to be a single vehicle without considering the force at the coupling. The UE regulation allows for two solutions, as shown in Fig. 9 [8].

Limits of adhesion utilization in accordance with Commission Delegated Regulation (EU) 2015/68 [8]: a – first solution, b – second solution
The first solution: the adhesion utilization rate for each axle group must satisfy the condition of ensuring the minimum required braking performance:
Second solution: the axle adhesion utilization rates should be within a certain range, then the wheel locking limits are established by the following relationships:
Furthermore, the adhesion curve for the rear axle assembly should satisfy the requirement:
For accurate calculations, the divisor in inequality (86) should be set to 0.7381.
The requirements concerning the wheel locking sequence are met if the adhesion utilized by the front axle is greater than that utilized by at least one of the rear axles at braking rates between 0.15 and 0.30 [8]:
Of course, in the case of the application of these solutions for a semi-trailer, the requirements for the front axle relate to the coupling device.
In the air braking systems of agricultural trailers, various types of load-dependent brake force regulators are used to approximate the ideal brake force distribution. The automatic load-sensing valves (LSVs) currently fitted to heavy trailers are designed to adjust the brake pressure on the axles according to the load condition [41]. If the braking forces are designed correctly, this will prevent the wheels from locking when the vehicle is unladen or only partially laden. On mechanically suspended trailers, the regulation is proportional to the spring deflection, while on air-suspended trailers, it depends on the control pressure of the air springs. If there are technical reasons against equipping the vehicles with an LSV (especially unsuspended vehicles), agricultural trailers or machines should be equipped with a manual brake force regulator. In the most popular three-stage adjustment device (full - half - empty), the regulation of braking force is achieved by pressure limitation in axle brake chambers [41]. Due to the difficulty to comply with the requirements of EU 2015/68 for the distribution of braking forces on vehicles with manual regulators, which were mentioned before, BPW developed a seven-position mechanical load-dependent brake force regulator (MBL), but with linear characteristics [42]. The BPW MLB works with a proportional pressure control. As a result, the output pressure remains proportional to the control pressure. This kind of control fully complies with the requirements of the EU regulation for unsuspended vehicles.
As the pressure distribution characteristic of the ALB and MLB is essentially a straight line, the distribution of braking forces between the coupling device and the rear axles can also be regarded as linear (radial). The contribution of the coupling device and trailer axles to the braking of the semi-trailer is expressed by the ratio of the partial braking force on the coupling device or individual axle to the total braking force:
The values of the braking force distribution coefficients defined in this way can theoretically vary from 0 to 1 and satisfy the following relationships:
Using the relations (88) and (89), the braking force of the coupling device and trailer axles can be calculated:
A directional coefficient of the brake distribution line, which passes through the origin of the coordinate system
Similarly, a linear braking force distribution, variable or fixed (in the absence of a braking force regulator), can be applied to the tandem assembly:
Unlike the β coefficients, the values of the
To find optimal solutions for the linear brake force distribution, the Quasi Monte Carlo method [43,44,45] was chosen. Fig. 10 shows an example block diagram of the algorithm for the optimum selection of the braking force distribution coefficients

A block diagram of an algorithm for the optimization of brake forces of a semi-trailer with tandem suspension using the Monte Carlo method (
The optimal values of the brake force distribution coefficients are determined by minimizing the objective function consisting of the residual sum of squares:
The OF thus obtained favourable solutions with the smallest differences between the adhesion utilised by each axle. As it is more important to reduce the difference between the values
Before the computation of the OF, the inequality conditions (83), (84) for the first solution and (85), (86) for the second solution are checked:
To simplify the notation of the boundary equations, they have been expressed as the product of the algebraic and logical equations. For the first solution:
For the second solution:
Then, the condition (87) for multi-axle trailer is checked:
In addition, an extra condition has been added to the rear axle adhesion utilisation rates:
Based on the algorithm described above, a computer program was written in MATLAB [46] to calculate the optimum distribution of the braking forces for semi-trailers with tandem suspension. The same algorithm, but without calculating the coefficients
The technical data and calculation results of the brake force distribution for the laden and unladen single axle semi-trailer are shown in Tab. 1.
The technical data and results of the optimization of brake force distribution in single axle semi-trailer:
UI-UII | 2250 | 3.895 | 0.98 | 0.0947–0.0947 | 0.2079–0.2079 | 3.8097–3.8097 |
LI - LII | 7250 | 3.840 | 1.25 | 0.1272–0.1272 | 0.2445–0.2445 | 3.0901–3.0901 |
In the case of laden and unladen semi-trailers, identical results were obtained for the optimum distribution of braking forces using both solutions. The adhesion utilization curves of the braked axle and the trailer coupling are shown in Fig. 11.

Adhesion utilization curves fi
The technical data of the laden and unladen semi-trailer with different tandem suspensions taken for the optimization calculations are presented in Tab. 2. To achieve comparability of the calculation results, the same mass
Rigid drawbar trailer and tandem suspension technical data [37–38]
unladen | laden | bogie (3.1) | 2 leaf spring (3.2) | 2 leaf 2 rod (3.3) | 2 leaf equal. (3.4) | air susp. (3.5) |
---|---|---|---|---|---|---|
m=3900 kg | m=19800 kg | d1=0.705 m | c1=0.454 m | c1=0.497 m | c1=0.454 m | c1=0.5 m |
L1=3.94 m | L1=3.94 m | d2=0.645 m | c=0.93 m | c=0.97 m | c=0.93 m | c=0.88 m |
L2=1.35 m | L2=1.35 m | hs=0.567 m | hs=0.717 m | hr1=hr1=0.467 m | hs=0.717 m | hs=0.717 m |
a=4.26 m | a=4.055 m | h2=0.547 m | h2=0.567 m | h2=0.567 m | h2=0.567 m | h2=0.567 m |
h=1.19 m | h=1.62 m | b2=0.03 m | d1=d2=0.21m | d1=d2=0.19 m | d1=d2=0.675 m | |
hh=0.59 m | hh=0.59 m |
The results of the optimization of brake force distribution in a tandem axle semi-trailer (L ‒laden, U ‒ unladen, Lw, Uw – laden and unladen with weight of suspension)
Bogie (3.1) | U-Uw | 0.3040–0.3115 | 0.1629–0.1588 | 0.5740–0.5793 | 0.2631–0.2619 | 5.1402–5.2971 | 0.4583–0.4521 |
I and II solution | L-Lw | 0.3006–0.3015 | 0.2301–0.2301 | 0.5286–0.5286 | 0.2412–0.2412 | 3.3456–3.3456 | 0.4564–0.4564 |
2 leaf 2 rod (3.3) | U-Uw | 0.6716–0.8831 | 0.2313–0.2307 | 0.1803–0.1527 | 0.5884–0.6166 | 3.3239–3.3354 | 3.2635–4.0377 |
I solution | L-Lw | 0.5841–0.6991 | 0.3117–0.3267 | 0.1674–0.1644 | 0.5209–0.5189 | 2.2086–2.1574 | 3.1116–3.1573 |
II solution | U-Uw | 0.9446–1.2169 | 0.2075–0.2065 | 0.1705–0.1444 | 0.6220–0.6490 | 3.8198–3.8417 | 3.6483–4.4935 |
L-Lw | 0.7531–0.7893 | 0.2806–0.2818 | 0.1577–0.1525 | 0.5617–0.5657 | 2.5635–2.5490 | 3.5631–3.7096 | |
2 leaf equal. (3.4) | U-Uw | 0.2512–0.2512 | 0.1951–0.1951 | 0.4031–0.4031 | 0.4018–0.4018 | 4.1265–4.1265 | 0.9969–0.9969 |
I solution | L-Lw | 0.2099–0.2099 | 0.2605–0.2605 | 0.3710–0.3710 | 0.3685–0.3685 | 2.8388–2.8388 | 0.9935–0.9935 |
II solution | U-Uw | 0.3002–0.3002 | 0.1803–0.1803 | 0.4040–0.4040 | 0.4156–0.4156 | 4.5449–4.5449 | 1.0288–1.0288 |
L-Lw | 0.2117–0.2117 | 0.2561–0.2561 | 0.3721–0.3721 | 0.3717–0.3717 | 2.9046–2.9046 | 0.9989–0.9989 | |
air susp. (3.5) | U-Uw | 0.2511–0.2511 | 0.1951–0.1951 | 0.4031–0.4031 | 0.4018–0.4018 | 4.1265–4.1265 | 0.9969–0.9969 |
I solution | L-Lw | 0.2098–0.2098 | 0.2605–0.2605 | 0.3710–0.3710 | 0.3685–0.3685 | 2.8388–2.8388 | 0.9935–0.9935 |
II solution | U-Uw | 0.3003–0.3003 | 0.1803–0.1803 | 0.4040–0.4040 | 0.4156–0.4156 | 4.5449–4.5449 | 1.0288–1.0288 |
L-Lw | 0.2117–0.2117 | 0.2561–0.2561 | 0.3721–0.3721 | 0.3717–0.3717 | 2.9046–2.9046 | 0.9989–0.9989 | |
2 leaf (3.2) only | U-Uw | 5.3013–3.5221 | 0.2271–0.2337 | 0.0017–0.0218 | 0.7712–0.7446 | 3.4025–3.2798 | 458.23–34.199 |
I solution | L-Lw | 4.6761–4.5855 | 0.3230–0.3202 | 0.0018–0.0030 | 0.6752–0.6768 | 2.0957–2.1233 | 376.30–222.04 |
Following the optimization criterion used (lowest OF values), the air suspension (section 3.5) and the two leaf spring suspensions with equalization (section 3.4) can be regarded as the best. The same optimum brake force distribution ratios were found for these tandem suspensions (Table 3) using the solutions described in Section 2. The values after the dash have been obtained by considering the weight of the tandem axle. Moreover, with equalised values of the

Adhesion utilization curves fi
Larger values of the OF were obtained for the bogie suspension (section 3.1) and even higher values for the two leaf-two rod suspension (section 3.3). In both cases, this causes a greater deviation of the adhesion utilization curves from the ideal distribution of braking forces (Fig. 12-b, d and Fig.13-a, c).

Adhesion utilization curves
For the two-leaf suspension (section 3.2), the calculation of the distribution of braking forces was only obtained from the first solution, with values of the OF objective function an order of magnitude higher (from 3.5 to 5.3) than for the other tandem suspensions. However, this solution cannot be considered correct either, as the leading tandem axle is braked to a very small extent (
A comparison of the results obtained without and with the weight of the tandem shows that for most tandems, the effect of this weight on the distribution of braking forces has been negligible. The differences in the brake force distribution ratios are no more than 5%. But for suspension 3.3, the differences in the calculated values of the
The method described for optimising the choice of linear brake force distribution for rigid drawbar agricultural trailers with single and tandem axles can be used in the initial design phase of air braking systems using the ALB or MLB regulator with radial characteristic. The calculation of the braking force distribution considered the requirements of EU Directive 2015/68 [8] for multi-axle trailers, treating the coupling device of a rigid drawbar trailer as a contractual front axle. It should be noted that although optimising the brake force distribution improves the braking performance and directional stability of agricultural vehicle combinations on different road surfaces, it does not prevent individual wheels from lockup, particularly on slippery and uneven surfaces. Therefore, agricultural vehicles travelling at speeds above 60 km/h must be equipped with ABS systems [8].
Optimization calculations using the Quasi Monte Carlo method for a rigid drawbar trailer with a payload of approximately 16 tonnes showed that the distribution of braking forces depends significantly on the type of tandem suspension. The lowest values of the minimized objective function were obtained for tandem axles with air suspension and two leaf springs suspension with equalization. For these two tandem assembly, the adhesion utilizations of the individual axles are closest to the straight line representing the ideal brake force distribution (Fig. 12-a,c), where the adhesion utilized by each tandem axle is the same and equal to the braking rate. Higher values of OF were found for the bogie suspension (Fig. 12-b, d), and even higher for the two leaf-two rod suspension (Fig. 13-a, c). In both cases, this leads to a greater deviation of the adhesion utilisation curves from the ideal distribution of the braking forces. The two leaf spring suspension (Figure 13-b, d) produced the highest OF values. These results are in qualitative agreement with Limpert's analysis of the braking dynamics of a vehicle combination with tandem axles [4]. In addition, calculations for the two leaf spring suspension have shown that the load transfer between the tandem axles can lead to premature locking of the leading axle wheels at a braking rate of about 0.6, which is also confirmed in the literature [3, 4]. The results of the optimisation calculations show that for most tandem suspensions, the effect of the suspension mass on the brake force distribution is negligible.