Document

advertisement
Wet Granulation
Scale-up Experiments
Scale-up Approach with Dimensional
Numbers
• The effect of process parameter (i.e., impeller speed, liquid addition rate) on the process
and thus on granule characteristics are evaluated based on mechanistic understanding.
• Use dimensional numbers as variables rather than individual process parameters. This will
decrease the number of variables need to be varied, thus the number of experiments (
includes explicitly includes liquid addition rate, implicitly includes impeller speed and liquid
addition mode).
• Determine the design space at the smallest scale for the given formulation based on
dimensional numbers and validate the design space for larger scales with less experiments
compared to DOE approach.
10
0
Mechanical
Dispersion
Crumb
10
-1
1
St
p
def
Intermediate
10
-2
Nucleation
Steady Growth
Rapid
Growth
0.1
10
-3
Drop
Controlled
Caking
10
0.01
0.1

1
a
Induction
-4
0
0.2
0.4
0.6
Smax
0.8
1
1.2
1.4
2
Design of Scale-up Experiments
Optimum conditions from Duquesne University Wet Granulation
experiments:
Liquid content
:5 %
Liquid addition rate :15 ml/min
Impeller Speed
:500 rpm
Wet massing time
:30 sec
The liquid percentage is kept same as the smallest scale
experiments. The sensitivity to liquid amount is tested around
the optimum liquid level that was obtained at small scale.
3
Design of Scale-up Experiments
• Ratio of fill height to granulator diameter at all
scales should be kept constant.
Granulator size
Granulation Batch Size (kg)
Diameter of the granulator
bowl (cm)
Fill height / Diameter
4l
0.6
17
10 l
1
24.6
75 l
9.72
52.5
0.27
0.15
0.15
4
Design of Scale-up Experiments
• The spraying time and the dimensionless spray flux
need to be kept constant.
4l
 = 0.11
For constant spraying time, the  values for larger scales with
single nozzle are calculated assuming 100 m drop size:
10 l
 = 0.15 and 0.18 (Close to small scale)
75 l
 = 0.34 and 0.50 (Too high compare to small
scale, therefore higher amounts of
lumps should be expected. Can be
5
lowered by using two nozzles)
Design of Scale-up Experiments
• Impeller speed was scaled according to constant tip speed
and constant Froude number (Fr) rules (provided that the
impeller speed is above the critical Froude number for
constant tip speed rule.
Constant Fr:
N 2 D1

N 1 D2
4l
Constant tip speed:
N: Impeller speed
D: Impeller diameter
10 l
N 2 D1

N 1 D2
75 l
340 rpm
500 rpm
195 rpm
420 rpm
290 rpm
6
Intermediate Scale (10 l)
Batch
96251-1
96251-2
96251-3
96251-4
96251-5
96251-6
96251-7
96251-8
96251-9
96251-10
96251-11
96251-12
96251-13
96251-14
96251-15
Impeller Liquid Content (%
Liquid
Liquid
Wet
Speed of liquid amount Addition Rate addition
Massing
(rpm)
to solid amount)
(g/min)
time (sec) Time (sec)
420
420
420
420
340
340
420
420
420
420
420
420
420
420
420
6
5
3
4
4
4
4
5
5
5
5
5
5
5
5
28.5
28.5
28.5
28.5
28.5
249.6
249.6
28.5
28.5
28.5
28.5
28.5
28.5
28.5
28.5
126
105
63
84
84
10
10
105
105
105
105
105
105
105
105
30
30
30
30
30
30
30
60
0
30
30
30
30
30
30
a
Median
Particle Size
(µm)
% of lumps (>
1 mm)
0.15
0.15
0.15
0.15
0.18
0.18
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
0.15
280
175
220
201
207
212
286
275
277
283
297
302
289
293
1.58
0.20
0.69
0.20
1.08
0.90
2.02
3.91
4.04
2.00
2.94
2.32
3.02
2.11
7
Effect of Liquid Content
80
3%
4%
5%
70
60
f(lnx)
50
40
30
20
10
0
2
10
3
Size (  m)
10
8
Summary
• Lump formation is low due to the low drop
penetration time and low dimensional spray flux.
• The granulation material is easy to process at low
liquid contents ( < 6%), not large clumps are
formed, but the whole mixture become very
sticky if the critical moisture content is exceeded.
• The formulation seem to be insensitive to most of
the operating conditions with the given fill ratio,
granulator geometry, and drying conditions and
analysis methods.
9
Large Scale (75 l)
Batch
Impeller Liquid Content
Speed
(% of liquid
(rpm) amount to solid
amount)
96251-21
290
5
Liquid
Addition
Rate
(g/min)
277
96251-22
290
5
277
96251-23
195
5
277
96251-24
290
5
277
96251-25
290
4
277
96251-27
195
5
277
Liquid
addition
time
(sec)
Wet
Massing
Time
(sec)
105
30
105
30
105
30
105
30
84
30
105
30
a
Median
Particle
Size
(µm)
% of
lumps
(>1
mm)
0.34
272
284
10.80
13.52
0.50
280
10.28
0.34
422
33.51
0.34
250
9.56
0.50
301
17.8
0.34
10
Comparison of PSDs- 10 l and 75 l
80
10 L
75 L- const. Fr
75 L- const. tip speed
70
60
f(lnx)
50
40
30
20
10
0
2
10
3
Size (  m)
10
11
Summary
• Although the median particle sizes matches
with the 10 l scale, the amount of lumps are
much higher at 75 l due to the doubled
dimensional spray flux.
• Further analysis of Stdef and Smax is needed to
evaluate the effects of liquid amount and
impeller speed on the PSD.
12
Download