Agitator Design for Gas-Liquid Fermenters and Bioreactors. Gregory T. Benz

Читать онлайн.
Название Agitator Design for Gas-Liquid Fermenters and Bioreactors
Автор произведения Gregory T. Benz
Жанр Химия
Серия
Издательство Химия
Год выпуска 0
isbn 9781119650539



Скачать книгу

3.4 Typical power number curve.

Graph depicts a typical pumping number curve.

      (3.14)equation

      The range between laminar and turbulent is called the transition range. It is a very broad range. That is because turbulence commences first near the impeller blade tips somewhere in the region of NRe = 10, but does not reach all of the far corners of the tank until NRe > 20 000. The power number gradually levels out in this range. For some impellers, it goes through a minimum in the transition range.

      In laminar flow, the pumping number becomes a constant. This means that under such conditions, the impeller pumping is independent of viscosity, though the power draw is directly proportional to viscosity.

      When vendors quote a pumping number, it is usually in turbulent flow and at a D/T of 1/3 and a C/T of 1/3. Many consider this geometry to be “standard”, but there is no fundamental reason to adhere to this geometry when designing agitation equipment.

      We can observe that blend time becomes constant in both laminar and turbulent flow. However, the laminar dimensionless blend time is often several orders of magnitude greater than the turbulent blend time. Specialized impellers have been developed for laminar flow mixing. We need not delve into these in this book, as fermenters never operate in the laminar range, because it is basically impossible to both incorporate gas into highly viscous liquids and have the gas exit in a reasonable amount of time after the gas is depleted.

      Some versions of the dimensionless blend time curve incorporate the D/T effect into the Y‐axis expression, as stated under the dimensionless blend time definition given earlier.

      Gassing factor depends on Aeration number, Froude number, D/T, Reynolds number, and impeller type. Therefore, it is impossible to show the entire relationship on a simple two dimensional graph. Figure 3.7 is based on turbulent flow, a D/T of 1/3, and a Froude number of 0.5.

Graph depicts the plot of dimensionless blend time. Graph depicts gassing factors.

      We will cover other dimensionless parameters, such as for heat transfer, as needed in subsequent chapters. For now, we will show some example calculations.

      Example 1: Power Draw Calculation

      A tank has an impeller of 1000 mm diameter, rotating at a shaft speed of 125 rpm, in a fluid with a specific gravity of 1.2. It has a known power number of 0.75. How much power will it draw?

      Answer

      (3.15)equation

      Rearranging to calculate power:

      (3.16)equation

      Since it is more common to think of power in kW, this would become 8.13 kW. An 11 kW motor (a standard motor rating) would be ample to power this impeller under these conditions.

      Example 2: Pumping Calculation

      A tank has a hydrofoil impeller of 6 ft. diameter rotating at 30 rpm. It has a known pumping number of 0.5. How much fluid will it pump?

      Answer

      (3.17)equation

      Rearranging to calculate Q (pumping rate), we get:

      (3.18)equation

      Had we used SI units, the calculated pumping capacity would have been expressed as m3/s, but the physical quantity represented would have been the same.

      Example 3: Blend Time Calculation

      This really only gets interesting in transition flow, where dimensionless blend time varies with Reynolds number. But let’s suppose we have an impeller with a dimensionless blend time of 10. How quickly would it blend the tank operating at 30 rpm?

      Answer

      If τ*N =10, τ = 10/N = 10/30(1/min) = 1/3 min = 20 s. Note that one can use any units for shaft speed; the resultant blend time units will be determined by this. Note also that as long as the dimensionless blend time is fixed, the result is independent of tank size. This means, for a given D/T and Reynolds number, the blend time at 30 rpm will be the same in any size tank. However, the P/V goes up exponentially when trying to keep the same D/T and the same shaft speed. So, designing for the same blend time in a large tank as a small one can be problematic.