Abstract
Compression speed is a critical parameter for the manufacturing of pharmaceutical tablets. Tableting scale-up often significantly increases compression speeds, frequently resulting in different final tablet characteristics or defects like capping or lamination. For this reason, compaction simulators are widely utilized to emulate the punch displacement cycles of production-scale rotary presses. While conventional simulator cycles rely on theoretical equations derived strictly from the geometric features of rotary equipment, they neglect structural frame compliance, which considerably alters the true velocity profile during tableting.
The present work introduces a novel, universally applicable analytical methodology to calculate the actual, deformation-compensated compression speed by accounting for the deformation of the press and the compressibility of the tablet formulation. The method was experimentally validated against empirical punch kinematics during the loading phase on an instrumented compaction simulator. The model was then applied to a production-scale rotary press, quantitatively demonstrating that structural machine deformation reduces the actual compression velocity to nearly half of the theoretical value near peak force.
The proposed method relies entirely on data readily determined in standard development and manufacturing environments, providing a highly practical tool for aligning compaction simulator profiles with production equipment, optimizing process scale-up, and troubleshooting compression speed-related production challenges.
Introduction
Tableting is one of the most important processes in the production of small molecule solid dosage forms. As with most other unit operations normally used for the manufacture of such dosage forms, the operational parameters employed during a tableting process play a key role in determining the final properties of the tablet such as its solid fraction, i.e. the relative density of the compact, strength (hardness), and defect-free appearance. Among these parameters, compression speed (i.e. the speed at which the powder is compressed) is particularly important. For example, it has been well established that, for the same compression pressure, the final density and tensile strength of the tablet depend on the compression speed–a trend generally referred to as “strain-rate sensitivity”.1 Furthermore, classical tablet defects such as capping or lamination are also strongly influenced by compression speed.2, 3, 4, 5, 6
The effect of compression speed on the aforementioned tablet qualities is of particular concern during the development of pharmaceutical formulations and processes because compression speed is often a direct consequence of scaling up the tableting process. Large commercial-scale rotary presses are equipped with dozens of stations—housings for the punch and die sets—that are arranged on the periphery of their large diameter turret. Since the linear velocity of the punch as it glides along the compression roller(s) determines the compression speed and is proportional to the turret radius, these large-scale production machines can generate much higher compression speeds compared to the small pilot- and laboratory-scale presses often used during the early development stages, which typically have significantly fewer stations and a smaller turret. Consequently, it is not uncommon for defects, that are absent during development and small-scale productions, to be observed only once the tablets are made on a large- or commercial-scale rotary press.
To bridge the gap between scales, compaction simulators are often used to empirically study the phenomena that can occur on large scale presses such as those introduced above. These machines, the first iterations of which were introduced as early as the 1970s,7 are essentially a single-station press that uses either a servo-hydraulic or an electromechanical system to drive the upper and/or lower punches to mimic the punch displacement observed on a given rotary press.8 This then allows tablet formulators to simulate the compression events occurring at large, commercial-scale presses using significantly less material than what would be needed to perform equivalent experiments at scale (i.e. on the actual presses).
The displacements prescribed to the movement of the compaction simulator punches are typically calculated using theoretical equations9,10 based on key geometric considerations that govern the compression event. They include the rotary press compression roller diameter, the punch head shape, and the turret diameter. Nevertheless, these equations do not account for additional factors that can be important when scaling up from small- to production-scale presses such as the deformation of the machine frame (also referred to as ‘frame stiffness’ in some previous work) and the punches.10
The concept of machine deformation has long been recognized in the literature and characterized for various rotary presses. For example, Altaf and Hoag measured the stiffness of a Stroke B2 rotary press and found its compliance to be 11.8 and 19.5 µm/kN for the upper and lower manifolds, respectively.11 This means that, when the press applies a compression force of 20 kN, the difference between the theoretical thickness, i.e. the vertical distance between the tip of the lower and upper punches that does not account for the press compliance, and the actual thickness at the corresponding force is 626 µm. Schmidt and Vogel12 performed the same measurement for a different rotary press, the Korsch PH 230, and found that that press deformed by the same order of magnitude as the Stroke B2 studied by Altaf and Hoag (around 650 µm at 20kN). Even though the deformation of the above two rotary presses was similar, additional studies showed that it can be machine-dependent, as demonstrated by Kennerley et al., who found differing compliance between the Manesty E2 and the Manesty F3 rotary presses.13
While press deformations measured in previous studies may seem small, they account for a significant portion of a typical tablet thickness (2-5 mm) at the point of maximum compression. As such, machine deformation can cause the punch movement on the actual press to significantly deviate from that determined from the theoretical equations.9,10 Such a deviation is particularly important in the case of compaction simulators since, as mentioned previously, these systems rely on the same theoretical equations to prescribe the movement of their punches. Consequently, studying the effect of compression speed using a compaction simulator may yield inaccurate conclusions.
To address this issue, Oates and Mitchell proposed a method to calculate the actual speed on a rotary press, taking machine deformation into account.14 This method was validated using various punches.15 Nevertheless, the proposed method is complex and difficult to use in practice as it requires adding a custom force instrumentation to the rotary press for measuring the force as a function of time throughout an entire compaction event. While this may be possible in some cases, it is often impractical or even impossible to implement on large commercial-scale presses that are typically used for production in GMP-regulated areas.
The present work proposes a simpler alternative method to calculate the actual speed of any rotary press. The method uses only the total machine deformation at different applied forces and the in-die compressibility of the formulation being compressed. In the next section, a detailed description of the proposed method is presented. Subsequently, we present a case study on how the method can be implemented on a compaction simulator equipped with displacement instrumentations. Finally, the application to a real rotary press will be demonstrated.
Download the full article as PDF here Quantifying the impact of press deformation on compression kinematics during tableting: An experimentally validated analytical approach
or continue reading here
Materials and Methods
Tablet presses
Experiments were performed using two different tableting equipment. The first was an electromechanical compaction simulator, STYL’One Evolution (MEDELPHARM, Beynost, France), set up with a set of Euro D round flat-faced punches with a diameter of 11.28 mm and their matching die. The actuator driving each punch was also equipped with displacement and force sensors.
To obtain the actual thickness during compression from the sensor data, a punch-on-punch experiment was performed to measure the difference between the thickness measured by the sensor and the actual thickness, which is equal to zero during this experiment, as a function of the applied force. This experiment provided the extent of machine deformation not seen by the sensor at different compression forces. Using this data, the actual thickness could be calculated by subtracting the amount of elastic deformation from the sensor-measured thickness using the force at the corresponding instant. The actual thickness is calculated automatically by the STYL’One software using the result from the punch-on-punch experiment.
When performing the punch-on-punch calibration, the result does not represent the machine’s total deformation. This is because the position sensor itself moves when force is applied, meaning the sensor’s raw reading already includes a portion of the machine’s flexing. Therefore, if the actual thickness is subtracted from the sensor’s thickness reading, only the ‘unseen’ deformation is captured. To determine the total machine deformation, the deformation actually recorded by the sensor (the difference between the sensor’s reading at zero force versus its reading under applied force) must also be added back.
The second tablet press equipment used for the experiments was a Korsch XL100 (Korsch, Berlin, Germany). This rotary press was set up with a set of domed TSM B round flat-faced punches with a diameter of 11.28 mm.
Powder
Experimental validation of the method described in Section 2 above was performed by compressing a blend comprising 99% w/w lactose monohydrate (Excipress, 150 Gr, Armor Pharma, Maen Roch, France) and 1% (w/w) magnesium stearate (Ligamed MF-2-V-MB, Peter Greven, Bad Münstereifel, Nordrhein Westafalen, Germany). Blend mixing was performed using a Turbula mixer (Type T2C, Willy A. Bachofen AG, Muttenz, Switzerland) for 5 min at 49 rpm.
V. Mazel, K.K. Sluga, E. Yost, N. Ho, L. Almeida, P. Tchoreloff, A.R. Muliadi, Quantifying the impact of press deformation on compression kinematics during tableting: An experimentally validated analytical approach,
Journal of Pharmaceutical Sciences, 2026, 104583, ISSN 0022-3549, https://doi.org/10.1016/j.xphs.2026.104583.
Read also our introduction article on Magnesium Stearate here:












































All4Nutra







