Investigation of a new mathematical model for compression of pharmaceutical powders

TY - JOUR

T1 - Investigation of a new mathematical model for compression of pharmaceutical powders

AU - Sonnergaard, Jørn

PY - 2001/1/1

Y1 - 2001/1/1

N2 - A new compaction equation, the log-exp model: V=V-w log(P)+V exp(P/P) is presented. The model presumes that two compaction processes: a logarithmic and an exponential decline may be active simultaneously. Using non-linear regression techniques the model gives an excellent fit to a number of model substances with wide differences in compaction behaviour. Compared to the Kawakita equation the model covers a broader range of the compaction profile. The new model and the Cooper and Eaton equation fit the data on the same level, but the parameters derived from the log-exp model seems to have more discriminative power between substances and have a close relation to the apparent plastic or brittle densification mechanism. The log-exp model has potential as a tool in estimation of the strength of agglomerated materials. A proposal for a set-up to an iterative non-linear regression calculation in a spreadsheet program is attached.

AB - A new compaction equation, the log-exp model: V=V-w log(P)+V exp(P/P) is presented. The model presumes that two compaction processes: a logarithmic and an exponential decline may be active simultaneously. Using non-linear regression techniques the model gives an excellent fit to a number of model substances with wide differences in compaction behaviour. Compared to the Kawakita equation the model covers a broader range of the compaction profile. The new model and the Cooper and Eaton equation fit the data on the same level, but the parameters derived from the log-exp model seems to have more discriminative power between substances and have a close relation to the apparent plastic or brittle densification mechanism. The log-exp model has potential as a tool in estimation of the strength of agglomerated materials. A proposal for a set-up to an iterative non-linear regression calculation in a spreadsheet program is attached.

UR - http://www.scopus.com/inward/record.url?scp=0034904364&partnerID=8YFLogxK

U2 - 10.1016/S0928-0987(01)00165-8

DO - 10.1016/S0928-0987(01)00165-8

M3 - Journal article

AN - SCOPUS:0034904364

SN - 0928-0987

VL - 14

SP - 149

EP - 157

JO - Norvegica Pharmaceutica Acta

JF - Norvegica Pharmaceutica Acta

IS - 2

ER -