CONSTRUCTION AND OPERATION OF AN IMPELLER RHEOMETER FOR ON-LINE RHEOLOGICAL CHARACTERIZATION OF NON-NEWTONIAN FERMENTATION BROTHS

CONSTRUCTION AND OPERATION OF AN IMPELLER RHEOMETER FOR ON-LINE RHEOLOGICAL CHARACTERIZATION OF NON-NEWTONIAN FERMENTATION BROTHS

A. C. BADINO JR. ¹ , M. C. R. FACCIOTTI ² and W. SCHMIDELL ²

¹Universidade Federal de São Carlos-Departamento de Engenharia Química (DEQ/UFSCar)

Caixa Postal 676 - CEP: 13565-905 - São Carlos - SP - Brazil

²Escola Politécnica da USP -Departamento de Engenharia Química (DEQ/EPUSP)

Caixa Postal 61548 - CEP: 05424-970 - São Paulo - S.P., Brazil,

Phone: [phone]; Fax: [phone]

(Received: June 11, 1997; Accepted: October 30, 1997)

Abstract - The present work describes the development and operation of a continuous rotational impeller rheometer for on-line rheological measurements in fermentation broths. In an initial step three devices were calibrated, two of them utilizing helical ribbon impellers and the third a six-bladed impeller. The calibration constants obtained were within the range mentioned in the literature, showing consistency with the methodology employed. In the sequence, the system with one of the helical ribbon impellers (HR39) was selected for on-line continuous rheological measurements due to its small dimensions and its wide measurement range. Finally, the performance of the proposed device, in terms of rheological measurements in Aspergillus awamori broths, was compared to that of a commercial bench rheometer. The comparison of on-line and off-line rheological measurements showed a good agreement between both measurement techniques. Furthermore, results suggest that the proposed on-line rheometer features higher sensitivity and easier operation than the commercial rheometer.

Keywords: Impeller rheometer, on-line rheological measurements, Aspergillus.

INTRODUCTION

Several economically important fermentation processes utilize filamentous microorganisms, generating broths very different from those of bacteria or yeast cultures. The complexity of mycelial broths results in highly viscous non-Newtonian fluids causing serious problems with momentum, heat and mass transfer. Therefore, it is important to determine the rheological properties of these broths throughout the fermentation.

The use of conventional bench rheometers, such as the concentric cylinder, to determine rheological properties of viscous mycelial suspensions, is often unsatisfactory. The main problems are caused by pellets’ size, generally of the same order of magnitude as the annulus, as well as the tendency of the suspension to become heterogeneous due to settling and particle interaction.

To circumvent these difficulties, Bongenaar et al. (1973) have proposed the use of a turbine impeller, instead of conventional impellers, to characterize mycelial fermentation broths. In fact, utilizing such an apparatus, Badino Jr. et al. (1994), have already characterized Penicillium chrysogenum and Cephalosporium acremonium broths and verified the influence of synthetic and complex media on the rheological behavior.

According to Reuss et al. (1982), the advantage of the turbine impeller system is the prevention of particle settling. However, since all measurements have to be carried out under laminar flow conditions, the range of shear rate involved is quite narrow and may not be relevant under actual fermentation conditions. Due to such limitations, these authors have proposed the alternative helical ribbon impeller for the rheological characterization of mycelial suspensions. Recently, Olsvik and Kristiansen (1994) concluded that on-line continuous measurement of broth rheology enables the uniform treatment of the samples and statistically more "correct" measurements.

The first on-line continuous rheological measurements of filamentous fermentation broths were done by Kemblowski et al. (1985). An experimental apparatus using a six-bladed impeller was proposed to characterize the rheological behavior of the Aureobasidium pullulans fermentation broth.

The experimental apparatus proposed by Kemblowski et al. (1985) was utilized by Olsvik and Kristiansen (1992) to study the influence of biomass concentration, specific growth rate and dissolved oxygen concentration on the rheological properties of Aspergillus niger fermentation broth.

The present work initially describes the construction and calibration of three rotational impeller rheometers and the procedure to choose the more adequate system for on-line continuous rheological measurements. Finally, it shows a comparison between the performance of the constructed apparatus and that of a commercial bench rheometer by considering experimental data from Aspergillus awamori batch cultivation.

MATERIALS AND METHODS

Calibration of the Rotational Impeller Rheometers

The rotational impeller rheometer consists of a cylindrical glass vessel with external water circulation to maintain constant temperature and a concentric stainless steel impeller. The impellers utilized for calibration, as well as the cylindrical glass vessel, are illustrated in Figure 1, and the dimensions and geometrical correlations are presented in Table 1. The impellers were designated as HR39, HR44 and 6BI, respectively.

The calibration theory was based upon the original reports of Bongenaar et al. (1973), slightly modified by Kemblowski and Kristiansen (1986) and Badino Jr. et al. (1994).

In order to determine the flow curve of a fluid by using a rotational impeller rheometer, the average shear rate, g _av, and the average shear stress, t _av, around the impeller at different rotational impeller speeds (N) must be determined.

According to Metzner and Otto (1957), the average shear rate, g _av, and the average shear stress, t _av, around the impeller should be determined as follows:

g _av = k N (1)

t _av = k’ T (2)

Figure 1: Rotational impeller rheometer - (a) helical ribbon impeller, (b) six-bladed impeller, (c) cylindrical glass vessel.

Helical ribbon 1 (HR39) Helical ribbon 2 (HR44) 6-bladed impeller (6BI) Di (m) H (m) L (m) Dt (m) HL (m) V (L) Dt/DiHL/H 0.039 0.039 0.0039 0.042 0.042 0.060 1.078 1.078 0.044 0.044 0.0044 0.048 0.048 0.090 1.091 1.091 0.044 0.044 0.0044 0.048 0.048 0.090 1.091 1.091

	Helical ribbon 1 (HR39)	Helical ribbon 2 (HR44)	6-bladed impeller (6BI)
Di (m) H (m) L (m) Dt (m) HL (m) V (L) Dt/Di HL/H	0.039 0.039 0.0039 0.042 0.042 0.060 1.078 1.078	0.044 0.044 0.0044 0.048 0.048 0.090 1.091 1.091	0.044 0.044 0.0044 0.048 0.048 0.090 1.091 1.091

Table 1: Rotational impeller rheometer - dimensions and geometrical correlations

where T is the impeller torque and k and k’ are calibration constants which depend exclusively on system geometry.

The calibration methodology utilized Newtonian fluids (silicone oils with densities, r , and viscosities, m , known for 25ºC) and non-Newtonian fluid (0.4% w/w xanthan gum aqueous solution).

Newtonian fluids were transferred to the cylindrical vessel in Figure 1.c and the temperature was maintained at 25ºC. Impeller torque (T) as a function of rotational impeller speed (N) was measured utilizing rheometers from BROOKFIELD Eng. Lab., Inc. (USA) (models LV-DVIII and RV-DVIII).

For a Newtonian fluid, the dependence of the power number, N_P, on the Reynolds number, Re, in the laminar flow region is given by equation 3, as follows:

(3)

where

(4)

(5)

(6)

(power consumption during agitation)

This procedure allowed the determination of the values of constant c (shape factor) which is dependent on the geometry of the systems.

In the sequence, the 0.4% w/w xanthan gum aqueous solution was characterized at 26ºC as a pseudoplastic fluid (, where K is the consistency index and n is the flow behavior index) utilizing a conventional concentric cylinder rheometer, BROOKFIELD model LV-DVIII. In addition, under the same temperature condition, torque values were obtained for the three investigated impellers, as a function of rotational speed. Thus, the average shear rate, g _av, could be calculated according to Badino Jr et al. (1994) by equation 7:

(7)

Average shear rate (g _av) and rotational impeller speed (N) values were plotted and the calibration constants, k, could be determined by linear regression of equation 1 for each impeller system. The other calibration constants, k’, were calculated by equation 8:

(8)

On-line Continuous Measurements

After calibration of the rotational impeller rheometers, an apparatus for the on-line continuous rheological measurements in fermentation broths was constructed (Figure 2).

Figure 2: Schematic diagram of the on-line rheometer apparatus.

Fermentation broth is pumped out of the bioreactor through a silicone rubber tubing, into the gas separator by a peristaltic pump. Large gas bubbles are removed from the liquid at the top of the separator and flow to the head of the cylindrical glass vessel. The fermentation broth accumulated at the bottom of the gas separator flows upwards through the rheometer. The liquid level is maintained constant by a second peristaltic pump that operates at a higher rate than the first one.

The outlet of the impeller shaft in the rheometer is provided with a rotating labyrinth seal. Aseptic conditions are ensured by introducing sterile air through a glass tubing situated beside the head of the vessel.

The shaft of the impeller is connected to a BROOKFIELD rheometer for measurements of the impeller torque at different rotational impeller speeds.

The average residence time in the loop was fixed at 90 seconds. This value allowed for good measurements without the occurrence of particle settling or even disturbance of cultivation performance.

In order to produce the broth to test the proposed on-line rheometer, Aspergillus awamori NRRL 3112 was cultivated in a medium containing cassava flour as the main carbon source in a 10 L bioreactor at 35ºC and a pH of 5.0. Specific aeration flow rate and agitation frequency were 0.5 vvm and 700 rpm, respectively. On-line and off-line rheological measurements were carried out every 3 hours starting 90 minutes after tank inoculation. Off-line rheological data were obtained with a BROOKFIELD model LVT bench rheometer, provided with a concentric cylinder system. On-line continuous rheological data were logged by a microcomputer coupled to the measuring head.

In on-line continuous measurements, it was possible to obtain average shear rates in the range of 0.5~40 s^-1 at the laminar flow region.

The power law model () was fitted to both on-line and off-line experimental data, which allowed the comparison of the estimated rheological properties (K and n).

RESULTS AND DISCUSSION

In the first step of the calibration, the shape factors, c, were determined by plotting the power number, N_P, as a function of the Reynolds number, Re, for the three impellers. Figure 3 illustrates the plots and Table 2 shows the shape factor values, as well as the correlation coefficients, obtained from the regressions.

In the sequence, the flow curve of the 0.4% w/w xanthan gum aqueous solution is presented in Figure 4. The rheological properties obtained from the power law model () were K=1.63 Pa.sⁿ and n=0.24, respectively.

Figure 3: N_Pvs Re plot in the region of laminar flow for the Newtonian calibration fluid.

Impeller c (-) R2 k (-) R2 k' (m-3) helical ribbon 1 (HR39) 290.8 0.9997 28.91 0.9997 10530 helical ribbon 2 (HR44) 267.6 0.99994 26.14 0.99990 7210 6-bladed impeller (6BI) 566.2 0.9997 45.80 0.9995 5770

Impeller	c (-)	R2	k (-)	R2	k' (m-3)
helical ribbon 1 (HR39)	290.8	0.9997	28.91	0.9997	10530
helical ribbon 2 (HR44)	267.6	0.99994	26.14	0.99990	7210
6-bladed impeller (6BI)	566.2	0.9997	45.80	0.9995	5770

Table 2: Calibration constants and correlation coefficients

Figure 4: Flow curve for 0.4% w/w xanthan gum aqueous solution at 26ºC.

The last step was to determine the constants k and k’ for the three impellers investigated by verifying the dependence of the torque impeller (T) on rotational impeller speed (N) in the laminar flow region by using the solution of xanthan gum at 26ºC.

The regression of rotational impeller speed, N, and g _av data, according to equation 1, allowed the determination of the first calibration constant k. Hence, the second constant, k’, was determined by utilizing equation 8.

Figure 5 illustrates the calibration curves and Table 2 shows the constant values obtained for the three impellers investigated, as well as the correlation coefficients (R²) calculated from the regressions. Constants obtained were within the range mentioned in the literature (Nagata, 1975; Allen and Robinson, 1991; Kemblowski and Kristiansen, 1986), showing the consistency of the employed methodology.

The HR39 impeller was selected as the system for on-line continuous rheological measurements. Besides utilizing the smallest sample volume, this system also allows for rheological measurements within a wider range of g _av and t _av values.

In the subsequent step, on-line and off-line rheological measurements were carried out utilizing Aspergillus awamori broths as the test fluid. Figure 6 illustrates the flow curves obtained via on-line continuous measurements. A good fit between the power law model and the experimental data was achieved.

Figure 7 shows the time course of the rheological parameters K and n for both on-line and off-line measurements. The results show a good agreement between K and n values obtained by both measurement techniques. Differences can be attributed to several factors. For instance, off-line measurements were carried out in a conventional bench concentric cylinder rheometer. Since this equipment has a narrow range of rotational speeds (N_max=30 rpm) to provide rheograms, it allows particle settling which leads to inaccurate measurements. Additionally, in many cases torque values were within the lowest 10% of the scale, which also introduces considerable error into the measurements. On the other hand, the ascendent flow pattern imposed by the on-line rotational impeller rheometer prevents particle settling.

Rheograms in Figure 6, as well as the behavior indicated in Figure 7, show that the technique developed is consistent and sensitive for the detection of variation in rheological parameters during cultivations. Interestingly, an additional feature in the on-line technique is its capability to accurately determine the instant in which the culture medium is carbon source depleted, provided, of course, that readings are taken more frequently. When the carbon source is exhausted, the K value decreases instantaneously, as can be observed in Figure 7.

Figure 5: Determination of calibration constant k for the impellers.

Figure 6: Flow curves obtained for Aspergillus awamori broths by using theon-line impeller rheometer.

Figure 7: Time course of K and n obtained through on-line and off-line measurements.

CONCLUSIONS

· The theoretical-experimental methodology utilized for the calibration of the rotational impeller rheometers enabled the accurate determination of calibration constants and the values were within the range mentioned in the literature. This indicates that any impeller could be utilized for on-line measurements.

· Rheological properties (K and n) obtained in Aspergillus awamori broths with both on-line and bench rheometers were rather close values. Differences in measurement would be due to the inherent limitations of the bench rheometer.

· The apparatus proposed for on-line rheological measurements showed a good performance, considering its easy operation as well as the accurate experimental data generated.

· The proposed technique could be an important tool for process monitoring and control of processes involving the cultivation of filamentous microorganisms.

ACKNOWLEDGEMENT

The authors wish to thank the "Secretaria da Ciência, Tecnologia e Desenvolvimento Econômico do Estado de São Paulo" for its financial support (Proc. SCTDE Nº 00775/93).

NOMENCLATURE

c Shape factor, -

D_i : Impeller diameter, m

D_t : Diameter of cylindrical glass vessel, m

H Impeller height, m

H_L Liquid height, m

k Calibration constant of equation 1, -

K Consistency index, Pa.sn

k’ Calibration constant of equation 2, m-3

L Ribbon or paddle width, m

n Flow behavior index, -

N Dotational impeller speed, s-1

N_P Dimensionless power number, -

P Power consumption, W

Re Dimensionless Reynolds number, -

R² Correlation coefficient, -

T Impeller torque, N.m

V Liquid volume, L

• Allen, D.G. and Robinson, C.W., The Prediction of Transport Parameters in Filamentous Fermentations Broths Based on Results Obtained in Pseudoplastic Polymer Solutions, The Can. J. of Chem. Eng. 69, 498 (1991).
• Badino JR., A.C.; Barboza, M.; Belluzzo, R.P. and Hokka, C.O., Caracterizaçăo Reológica e Verificaçăo da Influęncia da Massa Celular na Reologia de Caldos de Fermentaçăo com Penicillium chrysogenum e Cephalosporium acremonium, Revista Brasileira de Engenharia 11, Nş 1, 43 (1994).
• Bongenaar, J.J.T.M.; Kossen, N.W.F.; Metz, B. and Meijboom, F.W., A Method for Characterizing the Rheological Properties of Viscous Fermentation Broths, Biotechnol. Bioeng. 15, 201 (1973).
• Kemblowski, Z.; Kristiansen, B. and Ajayi, O., On-line Rheometer for Fermentation Liquids. Biotech. Lett. 7, Nş11, 803 (1985).
• Kemblowski, Z. and Kristiansen, B., Rheometry of Fermentation Liquids, Biotechnol. Bioeng. 28, 1474 (1986).
• Metzner, A.B. and Otto, R.E., Agitation of Non-Newtonian Fluids, AIChE J. 3, 3 (1957).
• NAGATA, S., Mixing Principles and Applications, Kodansha, Tokyo (1975).
• Olsvik, E.S. and Kristiansen, B., Influence of Oxygen Tension, Biomass Concentration, and Specific Growth Rate on the Rheological Properties of a Filamentous Fermentation Broth, Biotechnol. Bioeng. 40, 1293 (1992).
• Olsvik, E.S. and Kristiansen, B., Rheology of Filamentous Fermentations, Biotechnol. Adv. 12, 1 (1994).
• Reuss, M.; Debus, D. and Zoll, G., Rheological Properties of Fermentation Fluids, Chem. Eng. 381, 233 (1982).

Publication Dates

• Publication in this collection
  06 Oct 1998
• Date of issue
  Dec 1997

History

• Received
  11 June 1997
• Accepted
  30 Oct 1997