3-days Training Course

September 26-28, 2011

Table of contents

Tromp curves & Longitudinal sieving

presented by Dipl. Ing (FH) Dirk Pröpper

o Tromp curve
o Axial sieving

2 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

o Tromp curve
o Axial sieving

3 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Content

o General explanation
o Characterizing figures
o Preparation of tromp curves
o Example - Calculation of tromp curve
o Interpretation
o Conclusion

4 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

General explanation
What is a tromp curve?

A tromp curve is a diagram

which allows the
characterization of the
separator efficiency.

5 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

General explanation

Tromp curve describes the actual performance of a separator

o Efficiency
o Sharpness of cut
o Cut size
o Agglomeration
o Mechanical state

6 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

General explanation
Ideal Tromp curve

The ideal Tromp curve is a

step function
from 0 to 100 %.

7 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

General explanation
Real Tromp curve

A real separator shows a

graph not like a sieve. The
graph is a function of the
particle size (x) with a
minimum value called the
BYPASS and an inclined
line on the right side. The
slope of the inclined line is
an indication of the
sharpness of the
separation.

8 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

General explanation

To characterize the
performance of a separator,
three material samples
have to be taken from the
grinding circuit.

These samples have to be

analyzed by a laser
diffraction in order to get the
particle size distribution.

9 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Characterizing figures
Specific values which characterize the tromp curve are:

o Bypass value „t“

o Imperfection factor „I“
o Cut size „x50“
o Sharpness of separation acc. Eder „“

10 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Characterizing figures
Bypass value
o The lower the bypass,
the better the efficiency
of the separator.
o A high bypass is an
indication for over-
grinding and energy
wasting.

11 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Characterizing figures
Bypass value

o Based on normal OPC Cement (3000- 4000cm²/g):

• 1st generation separator: t ~ 30 – 60 %
• 2nd generation separator: t ~ 10 – 35 %
• 3rd generation separator: t < 10 %

o The bypass is more important than the sharpness of separation

(imperfection „I“ and „“ acc. to Eder).

12 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Characterizing figures
Imperfection

The imperfection will be

calculated as follows:

( x75  x25 )
I
2  x50
High efficiency separators
<0.35

13 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Characterizing figures
Cut size
o The cut size x50 marks
the partikel size where
the amount of fines and
coarse is equal.

14 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Characterizing figures
Sharpness of cut

The sharpness acc. to Eder

will be calculated as
follows:
x25
x
x75
High efficiency separators
>0.50

15 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Preparation of tromp curves

The given results from the particle size distribution will be used to
calculate the tromp curve.
In a first step the average recovery of weights [vf ; vg] has to be
calculated:

vf 
 Q ( x)   Q ( x)
a
100 %
g
vg 
Q f ( x)  Qa ( x)
100 %
 Q ( xf)   Q ( x ) g Q f ( x )   Qg ( x )

With:
vf = recovery of separator fines as function from particle size;
vg = recovery of separator rejects as function from particle size;
Qa(x) = amount of separator feed passing [m-%];
Qf(x) = amount of separator fines passing [m-%];
Qg(x) = amount of separator reject passing [m-%];

16 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Preparation of tromp curves

The circulating load [u] can also be calculated from the measurement
results:

u
1

 Q ( x)   Q ( x)
f g

vf  Q ( x)   Q ( x)
a g

With:
vf = recovery of separator fines as function from particle size;
Qa(x) = amount of separator feed passing [m-%];
Qf(x) = amount of separator fines passing [m-%];
Qg(x) = amount of separator reject passing [m-%];

17 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Preparation of tromp curves

Further on, the size selectivity [T(x)] of a separator can be calculated

as follows:

Qg ( x)
T ( x)  v g 
Qa ( x)

NOTICE: This figure is a function of the particle size. This means

that this figure has to be calculated for each particle size analyzed in
the laboratory!
With:
ΔQa(x) = feed material amount [m-%] passing the screen size/width x;
ΔQg(x) = reject material amount [m-%] passing the screen size/width x;
vg = recovery of separator rejects as function from particle size;

18 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Example - Calculation of tromp curve

Particle size distribution from laboratory

Particle size xi Q3,Feed Q3,Fines Q3,Coarse

[μm] [%] [%] [%]
1.8 4.7 9.6 1.9
2.2 5.9 12.1 2.4
8.6 17.2 37.2 4.9
10 18.9 41.0 5.1
18 27.3 59.3 6.4
21 30.0 64.9 7.0
72 72.8 99.2 52.9
86 79.9 99.8 63.8
146 95.2 100 88.0
174 98.1 100 93.3
19 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Example - Calculation of tromp curve

Calculation of the sum of the passing material

Q3,Feed Q3,Fines Q3,Coarse

[%] [%] [%]
4.7 9.6 1.9
5.9 12.1 2.4
17.2 37.2 4.9
18.9 41.0 5.1
27.3 59.3 6.4
30.0 64.9 7.0
72.8 99.2 52.9
79.9 99.8 63.8
95.2 100 88.0
98.1 100 93.3
∑Qa(x) ∑Qf(x) ∑Qg(x)
450.1 623.2 325.7
20 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Example - Calculation of tromp curve

Calculation of the average recovery factors

vf 
 Q ( x)   Q ( x)
a g
100% 
450.1  325.7
100%  41.82%
 Q ( x)   Q ( x)
f g 623.2  325.7

vg 
 Q ( x)  Q ( x)
f a
100% 
623.2  450.1
100%  58.18%
 Q ( x )  Q ( x )
f g 623.2  325.7

21 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Example - Calculation of tromp curve

Calculation of average particle size of the fractions

xi Δx in µm mean particle size

[μm] [μm] [μm]
1.8    
0.4 2
2.2
6.4 5.4
8.6
1.4 9.3
10
8 14
18
3 19.5
21
51 46.5
72
14 79
86
60 116
146
28 160
174    
22 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Example - Calculation of tromp curve

Calculation of the fractional amount of the fines material

Qf(x) ΔQf(x) ΔQf(x)*vf/100

[%] [%] [%]
9.6
2.5 1.0
12.1
25.1 10.5
37.2
3.8 1.6
41.0
18.3 7.6
59.3
5.6 2.3
64.9
34.3 14.3
99.2
0.6 0.3
99.8
0.2 0.1
100
0 0
100
23 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Example - Calculation of tromp curve

Calculation of the fractional amount of the coarse material

Qg(x) ΔQg(x) ΔQg(x)*vg/100

[%] [%] [%]
1.9
0.5 0.3
2.4
2.5 1.5
4.9
0.2 0.1
5.1
1.2 0.7
6.4
0.6 0.4
7.0
45.9 26.7
52.9
10.9 6.3
63.8
24.2 14.1
88.0
5.3 3.1
93.3
24 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Example - Calculation of tromp curve

Calculation of the fractional amount of material in the feed

ΔQf(x)*vf/100 ΔQg(x)*vg/100 ΔQa(x) calc.

[%] [%] [%]
1.0 0.3 1.3
10.5 1.5 12.0
1.6 0.1 1.7
7.6 0.7 8.4
2.3 0.4 2.7
14.3 26.7 41.1
0.3 6.3 6.6
0.1 14.1 14.1
0 3.1 3.1
25 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Example - Calculation of tromp curve

Calculation of the Tromp curve values T(x)

ΔQg(x)*vg ΔQa(x) calc. T(x)ag=vg*ΔQg(x)/ΔQa(x)

26.8 1.3 20.2
147.9 12.0 12.4
11.5 1.7 6.7
71.0 8.4 8.5
37.0 2.7 13.6
2672.6 41.1 65.1
632.3 6.6 95.9
1406.9 14.1 99.5
310.6 3.1 100
26 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Example - Calculation of tromp curve

The diagram shows the calculated size selectivity
o Bypass 6.9%
o Imperfection
(46.0  24.3)
I  0.32
2  33.5
o Cut sharpness acc. to
Eder
24.3
x  0.53
46.0
o Cut size

x50  33.5

27 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Interpretation
Reasons for separator inefficiency

o Overloaded with material e.g. to high circulating load,

under-dimensioned
o Poor feed distribution on distribution plate
o Poor air distribution in the separating zone
o Separator operating point – air volume, rotor speed
o Overpressure

28 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Interpretation
Tromp curve effects

29 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Tromp curve

Conclusion

o Elaboration of the tromp curve allows characterization of the

potential in the separator
o Proper interpretation of the tromp curve allows early problem
definition e.g. dynamic seal worn-out
o The solution for each separator is not equal and engineering
experience is required to devise the appropriate solution
o The tromp curve analysis is completed by
• Material specific loads – feed and fines
• Operating point – air/rotor speed
• Process parameters

30 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

o Tromp curve
o Axial sieving

31 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Content

o Main reason for an axial sieving

o Information required for interpretation
o Axial sampling after crash stop
o Sieving and evaluation of results
o Coarse grinding
o Fine grinding
o Conclusion

32 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Main reason for an axial sieving

Why is it important to do an axial sampling

o Evaluation of the grinding progress in the mill

o Adaption of ball charge composition acc. to the cement type
o Visual analyse of the situation in the mill
o Try to get a feeling about the condition of the grinding plant

33 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Information required for interpretation (1/3)

o Fresh feed:
• Composition (cement type)
• Size distribution of the fresh feed
• Hardness of the fresh feed
o Chamber length / clear diameter
o Ball charge
• Ball charge composition
• Ball charge classification (specially 2nd chamber)

34 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Information required for interpretation (2/3)

o Liner shape and actual condition (specially 1st chamber)

o Intermediate diaphragm (slot opening, centre openings,
adjustments flow control)
o Mill speed
o Production rate and circulating load under normal stable
conditions (trends) before the crash stop
o Mill motor absorbed power
o Process paramter

35 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Information required for interpretation (3/3)

o Material level in the chambers

o Axial sieving results composed of:
• Cumulated residue on sieves ranging from:
• 16 – 0.5 [mm] in the 1st chamber
• 2 – 0.045 [mm] in the 2nd chamber
• Blaine values of samples in the 2nd chamber

36 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Axial sampling after crash stop

37 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Sieving and evaluation of results

Standard sheet (1/3)

38 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Sieving and evaluation of results

Standard sheet (2/3)

39 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Sieving and evaluation of results

Standard sheet (3/3)

40 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Coarse grinding

41 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Coarse grinding
1st chamber crushing efficiency

42 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Coarse grinding
Parameters you can change to increase the efficiency:

o Ball charge composition

o Ball charge filling level
o Diaphragm flow control adjustment
o Liner condition and design

43 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Coarse grinding
Ball charge composition

Ball charge too fine:

o Arise nibs because of less crushing effect
o Material level too high
o Retention time too high

44 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Coarse grinding
Ball charge composition

Ball charge too coarse:

o Avoid nibs
o Material too coarse due to the high void volume
o Material level too low Can be compensated
o Retention time too low with flow controlled diaphragm

45 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Coarse grinding
Standard ball charge for the 1st chamber

Add only Ø 90 &/or Ø 80 [mm]

balls to coarsen the ball charge

Add only Ø 60 &/or Ø 70 [mm]

balls to get a finer ball charge

Details:
Spec. ball surface: 10.58 m²/t
Avg. ball weight : 1,458 g/pcs.
Avg. ball-Ø : Ø 71.4 mm

46 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Fine grinding

47 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Fine grinding
2nd chamber grinding efficiency

48 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Fine grinding
Parameters you can change to increase the efficiency:
o Ball charge composition
o Diaphragm flow control
adjustment
o Ball charge filling level
o Liner condition and design
o Check gaps between
grates (if any should not
be bigger than grates slot
opening)
o Such situations should
be avoided!

49 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Fine grinding
Ball charge composition

o The feed material to the 2nd chamber and the cement type
decide the charge composition
o For the most applications a standard ball charge is
recommended.

o Some points to consider

• Ø 18 [mm] balls might not work well in case of cements with
high limestone content (> 5 – 10 [%])
• very fine cements (> 4’000 [cm²/g]) or with high amounts of
slag or puzzolana might benefit from finer charges than the
standard.

50 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Fine grinding
Standard ball charge for the 2nd chamber

Go without Ø 60 mm by proper
conditions in chamber 1

Go finer by grinding higher

fineness or composite cement
types

Details:
Spec. ball surface: 28.77 m²/t
Avg. ball weight : 51.3 g/pcs.
Avg. ball-Ø : Ø 23.4 mm

51 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Fine grinding
Standard ball charge for the 2nd chamber

Go without Ø 40 mm by proper
conditions in chamber 1

Go finer by grinding higher

fineness or composite cement
types

Details:
Spec. ball surface: 31.70 m²/t
Avg. ball weight : 51.3 g/pcs.
Avg. ball-Ø : Ø 23.4 mm

52 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Axial sieving

Conclusion

o Correct sampling under controlled conditions is very important!

o The 1st compartment performance is always the priority –

actions must be taken to improve the 1st compartment
performance before changing the 2nd compartment

o Improvement of the grinding performance should be a

progressive task – the more sievings you have, the more
confident you can be with the actions taken

53 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH
Any questions?
Feel free to ask!

Thank you for your

kind attention.

54 Tromp curves & Longitudinal sieving – September 26-28, 2011 © Christian Pfeiffer Maschinenfabrik GmbH