PressterÔ, Punch Displacement Profiles, and
Related Issues
Technical
Document No. 18
Last Revision: March 18, 2000
LVDT Calibration and Precision
Compaction Simulation User Group
The theoretical path of a tablet press punch is
calculated from the following equation introduced by Rippie & Danielson
(1981) and utilized by every compaction simulator:
z
= [(r1
+ r2)2
– (r3
* sin wt
– x)
2]½
where
z is the vertical displacement of the upper punch at time t
r1
is the radius of the compression roll
r2
is
the radius of the vertical curvature of the punch head rim
r3
is
the radius of the “pitch circle” (distance between turret and punch axes)
w is the turret angular velocity
There are several reasons why this
equation, although widely used to govern compaction simulator punch movement, is
inadequate for press replication:
The equation, as used in simulators, does not account for the flat portion of the punch head (Yang et al, 1986)
The equation describes the punch movement in and out of an empty die. It does not account for the elastic expansion of the tablet during unloading as the punch moves out of the die and the tablet is actually exerting a force on the punch.
On a typical rotary press, the lower punch displacement curve is reaching its peak before the upper punch profile does. The actual lower punch peak occurs well after the theoretical sinusoid curve.
The upper punch displacement profile depends on the force of compaction but the equation does not factor in the effect of pressure.
The equation does not take into account any deformation of the punch or the press parts under stress (Ruegger, 1996).
In addition to the geometry of the punch
movement, compaction simulators cannot reproduce the actual speed of the punches
in a press. Before the minimum of
punch penetration into the die is reached, the punch speed is slower than
theoretically predicted as the load on the driving motor increases causing it to
slow down. After that minimum, for
some cases, theoretical speed is less than the actual due to the elastic
expansion of the tablet. A huge
error results when the theoretical values of punch velocity are used for
calculation of power expenditure in tableting (Armstrong
and Palfrey, 1987).
The magnitude of the deviation of the actual punch movement from theoretical predictions depends on the applied force, press speed, material being compressed, and motor capacity. “It is therefore follows that if simulators are to be used to study the compaction process, the pattern of punch movement fed into the simulators must be adjusted to take these factors into account. Feeding in a uniform pattern of punch movement which is to be used under all circumstances may give rise to misleading results” (ibid).
Several
vendors (Puuman Oy, SMI) offer instrumented punch, that is, a punch that has
strain gages and other instrumentation built-in.
The data is then accumulated or transmitted via telemetry to a stationary
computer (Ilkka, 1998).
Such
devices are versatile enough to report compression forces and either punch
displacement or acceleration, and, at least in theory, they can be easily
moved from press to press.
However,
one should keep in mind that each instrumented punch is limited to one size
and shape of the tooling, and is limited to one station, compared to roll
pin instrumentation that reports data for all stations and any tooling. The
replaceable punch tip is not a reliable option as these tips so often break
and/or distort the signal (Houghton, 1999).
Puuman instrumented punches are rather cumbersome to install or move from press to press. They require two adjacent stations to be occupied by battery and LVDT conditioning unit. The precision of the punch displacement measurement is rather poor (110 microns). In addition, the measurement is affected by tilting of the punches (Matz et al., 1999).
SMI punches, on the other hand, are relatively easy to install but they report a useless measurement of punch acceleration instead of punch displacement. Acceleration variable cannot be integrated to produce a true punch displacement signal because the integration constants (zero-point velocity) are not known (Sirihorachai, 1999). Attempts to calculate displacement from acceleration so far could not be validated.
PressterÔ can provide reliable punch displacement profiles for a majority of production presses currently on the market. It might be interesting to compare the PressterÔ profiles with those obtained from instrumented punches.
Machine deformation leads to errors in the determination of punch displacement (Altaf and Hoag, 1995) and contact time.
Schmidt and Vogel (1993) on Korsch PH230 established that the deformation slope was in the range of 0.0244 mm/kN above 2.5 kN load and 0.0249 mm/kN above 0.64 kN. Oats and Mitchell (1989) observed on a Manesty Betapress two linear regions with a slope changing at approx. 2.3kN; the upper punch gave 35% of the total deflection, the lower punch gave 65%. Altaf and Hoag (1995) characterized Stokes B2 deformation as 0.012 mm/kN and 0.019 mm/kN for the upper and lower assemblies, respectively. For at least two simulators the total deformation (following the initial curved portion of the force-deformation graph) for F10 tooling was of the order of 0.010 mm/kN (Ruegger, 1996).
As per the results of the recent Presster testing, the deformation slope was evaluated to be about 0.018 mm/kN for both the upper and lower assembly with loads up to 40 kN.
In
a rotary press, punches move both vertically and horizontally during the
compaction of a tablet. When the
punch is in contact with the roll, any increment of horizontal motion (e.g. in
terms of the angle of the turret rotation) causes a proper vertical
displacement. This relationship
depends on the press and punch geometry and is described by Rippie and Danielson
equations.
If
we consider equal time intervals at constant horizontal speed, every increment
of the horizontal motion will be the same.
The vertical displacements and the vertical speed will change from some
maximum value at the first contact with the roll (the beginning of the
consolidation stage), to zero when the flat surface of the punch head passes the
roll (dwell time stage). Thus, the
punch position on the roll (as well as the roll diameter) defines the vertical
speed value.
The
maximum vertical speed value (for a given horizontal speed) will occur in the
case when the maximum punch travel during the consolidation stage is required.
The value of the max travel can be derived from the following
considerations.
For
the most production presses the max depth of fill is 21mm; max tablet thickness
is 10mm. In this case the max
displacement of each (upper and lower) punch to compress a tablet is
(21-10)/2=5.5mm. Using geometrical
calculations, it can be found that the max vertical speed at this position of
the punch on a Manesty Unipress Diamond is 561mm/s. As both punches move simultaneously, the max compression rate
is 561x2=1122mm/s.
Manesty Betapress features max
depth of fill equal to 17.4mm and max tablet thickness 8 mm.
The max displacement of a punch is (17.4-8)/2=4.7mm.
In this case the max vertical speed is 441mm/s, max compression rate is
882 mm/s.
The fastest compaction simulators
can maintain max compression rate of 3000mm/s.
These velocities may be of use for basic studies of a powder behavior but
are excessive for tablet press simulation.
The Presster is designed to mimic
the most presses in the industry. Its
max horizontal speed of 2200 mm/s can produce max compression rate of 1256 mm/s
for the Unipress geometry and of 1490 mm/s for the Betapress geometry.
The motion of punches in the horizontal plane is rotational on presses and linear on the Presster. If we run the Presster with a linear speed equal to the press rotational speed, the vertical speed of the punches will be slightly different. This difference changes from the max value at the first contact with the roll to zero at the end of the consolidation stage. For a Unipress Diamond, for example, the average speed difference during the consolidation stage can reach 1.65% for max tablet thickness of 10mm. There will be the same difference if we simulate on the Presster the compaction of an 8mm thick tablet on a Betapress. But for a more usual tablet thickness of 4mm, the difference drops down to 0.76% for a Betapress and to 0.68% for a Unipress.
From the results of Validation Testing of the Presster machine:
1. Without load the machine maintained the speed in the range of +/- 1%. The max horizontal speed of 2.18m/s (1800 RPM) was reached during the test.
2.
Under load the load of 25kN (50% of the max. load) the machine could
maintain 90% of the speed without load (at the same RPM) in the range of linear
speed of up to 1.5m/s (1400 RPM). The
max linear speed, which was reached reproducibly at the load of 25kN, is 1.73m/s
at 1600 RPM. The max force that was reached during the test was 31kN at 1.49m/s
(1450RPM).
3.
The punches hit the rolls heavily
at the max speed. Since the test
was conducted with small rolls (7” DIA), larger rolls may allow increasing the
max speed at 25kN load.
LVDT
Calibration and Precision
There are a number of issues concerning the accuracy of LVDT measurements on compaction simulators – see, for example, Juslin and Paronen (1980), Lloyd, York and Cook (1991), Ho et al. (1979), Holman and Marshall (1993). The reason for this concern is that the punch movement has to be closely monitored for any feedback to the hydraulic actuators under the displacement control operation.
The Compaction Simulator User Group (CSUG) has set a strict requirement regarding the non-linearity of displacement transducer. After much discussion, it was decided to limit any deviation from linearity to 15 µm for each transducer so that the total maximum error would not exceed 30 microns (CSUG, July 3, 1996).
The Presster, however, for its operation does not need the LVDT measurement at all. The punch displacement profile of a rotary press is matched by virtue of geometric similarity. Therefore, a pilot machine design did not call for high precision LVDT measurements.
During the Validation testing of the Presster, we compressed metal tablets with different thickness at different speeds and levels of applied pressures. The LVDT readings indicating the gap between punch tips were compared to the actual tablet thickness as measured by a precision caliper. The data was fitted into a regression equation that used compression force and speed to minimize the measurement error and to predict in-die tablet thickness. When the corrections are applied to data, the absolute gap measurement error is reduced to about 30 mm.
Compaction
Simulation User Group (CSUG)
In 1996, there were several recorded meetings of CSUG. Several reports are available on the net (CSUG March 23, July 3, October 2, 1996). Since then the group ceased to function.
Owing to a renewed interest in the compaction simulation and press replication in the U.S. and elsewhere, the time has arrived, perhaps, to re-organize and reconvene the group.
During the first of the original CSUG meetings, it was decided that “any machine which could attain certain minimum speeds and frequency responses could reasonably be described as a simulator”. If so, then the MCC PressterÔ certainly qualifies. MCC would be willing to lead the effort of getting a new CSUG on the cyberspace where virtual meetings could be held without the burden of physical travel.
Altaf
SA, Hoag SW. Deformation of the Stokes B2 rotary tablet press:
Quantitation and influence on tablet compaction. J Pharm Sci
84(3):337-343, 1995
Armstrong
NA, Palfrey LP. Punch velocities during the compaction process. J Pharm
Pharmacol 39:497-501, 1987
CSUG.
Compaction Simulator Users Group Meeting -
Standards http://www.bath.ac.uk/ March 23rd, 1996
CSUG.
Compaction Simulator Users Group Meeting -
Standards http://www.bath.ac.uk/ July 3, 1996
CSUG.
Minutes of the Compaction Simulator Users' Group http://www.bath.ac.uk/ October
2, 1996
Ho
AYK, et al. A comparison of three methods of mounting a linear variable
displacement transducer on an instrumented tablet machine. J Pharm Pharmacol
31:471-472, 1979
Holman
LE, Marshall K. Calibration of a compaction simulator for the measurement of
tablet thickness during compression. Pharm Res 10:816-822, 1993
Houghton R. Utility of an instrumented tablet punch. Arden House Conference, 1999.
Ilkka J. Instrumentation of Rotary Tablet Machine by a Portable Measuring System. In: Munoz-Ruiz A, Vromans H (eds.) Data Acquisition and Measurement Techniques. Interpharm Press, Buffalo Groves, Illinois, 1998, p. 239-255
Juslin
MJ, Paronen TP. On the accuracy of displacement measurements by instrumented
single punch machine. J Pharm Pharmacol 32:796-798, 1980
Lloyd
J, York P, Cook GD. Punch elasticity compensation in the calibration of
displacement measurements on a compaction simulator. J Pharm Pharmacol 43S:80P,
1991
Matz C, et al. On the accuracy of a new displacement instrumentation for rotary tablet presses. Drug Dev Ind Pharm 25:117-130, 1999
Oates
RJ, Mitchell AG. Calculation of punch displacement and work of compaction on a
rotary tablet press. J Pharm Pharmacol 41:517-523, 1989
Rippie EG, Danielson DW. Viscoelastic stress/strain behavior of pharmaceutical tablets: Analysis during unloading and postcompression periods. J Pharm Sci, 70:476-482, 1981
Ruegger CD. An Investigation of the Effect of Compaction Profiles on the Tableting Properties of Pharmaceutical Materials. Ph. D. Thesis , 1996
Sirihorachai, R. Use of a novel data acquisition system for determining punch displacement and forces on a rotary tablet press. Arden House Conference, 1999.
Yang L, Venkatesh G, Fassihi R. Characterization of compactability and compressibility of poly(ethylene oxide) polymers for modified release application by compaction simulator. J Pharm Sci 85:1085-1090, 1985