Open Source Bioreactor: Baseline and knobs
What this reactor is built to show now, what it holds at a standard so that can be shown, and which settings later work is meant to move. The reasoning behind each number lives in the document it links to; this page is the map.
What the first paper claims
Section titled “What the first paper claims”The contribution is the design: a stirred-tank photobioreactor that is parametric, derives its numbers from cited data, prints on a desktop printer and costs little. To make that claim the reactor has to be comparable with the literature, so wherever a standard configuration exists, the reference build runs at it. A choice of this project’s own would make every result a result about that choice.
Later papers turn the knobs: the impeller pair, the layout, the speed, the aeration. The model already carries them, and each is one parameter away. Nothing on this page is a claim that the standard is the best setting; only that it is the one a reader can compare against.
How the model holds a standard
Section titled “How the model holds a standard”- A knob defaults to its standard. Where a component carries one, the layout comes from the
component’s own row: the impeller types in
scad/custom/impellers.scadeach carry the tank layout they are standard at, andimpeller_bore_ratio,impeller_clearance_factorandimpeller_spacing_factorareundefuntil set. - A departure is reported, not hidden. A knob set to a number is echoed as a departure from the
standard, with both values (
impeller layout: D/T 0.45 departs from the standard's 0.333333). The echo transcripts intests/echorecord every build’s stream, so a departure shows in review. - Every number carries its kind: measured, correlated, borrowed from a source, or reasoned and not cited (Design conventions). A standard is a borrowed number, and the row says from where.
Agitation: the baseline
Section titled “Agitation: the baseline”On jar_10L_220x305 (bore T = 210 mm, culture H/T 1.09).
| knob | baseline | source | kind | what later work varies |
|---|---|---|---|---|
| impeller type | six-blade Rushton turbine at both positions | the standard fermentation configuration, on a microalgae researcher’s advice | chosen to be standard | a Rushton under a pitched blade (future work); a pitched pair |
| blade proportions | blades D/5 high, D/4 long, on a 3D/4 disc | McCabe 1993 p. 243 for the blades; Middleton & Ramshaw, US 5198156 for the disc | borrowed | none planned |
| power number | 4.17 +/- 0.14 | Kaiser 2016; Agitation says why not McCabe’s 5.75 | measured, by others | a bench measurement of the printed part |
| flow number | 0.65 | Grenville 2017, typical | borrowed | as above |
| D/T | 1/3 (70 mm) | the standard turbine, McCabe, Smith & Harriott 1993 pp. 242-243, after Rushton, Costich & Everett 1950; the Handbook of Industrial Mixing allows T/4 to T/2 | borrowed | 0.45, where the pitched pair ran |
| lower clearance | T/3 (1 D, 70 mm) | McCabe 1993: E/T 1/3 for one turbine, and about one impeller diameter off the bottom for several (p. 241) | borrowed | T/4, Rutherford et al.’s parallel flow |
| spacing | T/3 (1 D, 70 mm) | Rutherford et al. 1996’s merging flow, equal thirds at H = T, relayed by Li et al. 2007 Fig. 1; McCabe 1993 p. 253 asks at least one impeller diameter | borrowed | T/2, the parallel flow; 1.5-2 D |
| impeller count | 2 at H/T 1.09 | the microalgae researcher advising the project; McCabe 1993 p. 241 keeps one impeller to a depth equal to or somewhat over T and adds more for deeper | chosen to be standard; Rutherford’s pairs ran at H = T | a single Rushton |
| baffles | four at T/12 on the wall, T/72 off the glass, floor to surface: 1.00 of Oldshue’s reference area | McCabe 1993 p. 240 and p. 243, four at T/12; the Handbook gives T/12 to T/10; the gap is Dynamix’s J/6 | borrowed | T/10 blades; lid plates, which other builds hang |
What the baseline costs. The motor was chosen when the impeller was 94.5 mm, to reach both ends of the speed band Agitation builds from Leupold et al. 2012. On the 70 mm impeller its 320-420 rpm gives 1.17-1.54 m/s at the tip: the Chlorella optimum of 1.26 m/s falls at about 344 rpm, inside the band, and the 2.03 m/s break-even is out of reach. That band carries its own stated assumption, that a plate-and-skirt rig’s rim speed transfers to an impeller’s tip.
What it changes, on jar_10L at the rated 320 rpm: 1.06 W, 130 W/m^3 mean and 27.1 W/kg peak, the pair under 0.063 N-m of the 0.490 N-m rating. The gas holds at 0.5 vvm down to 242 rpm.
Other jars. The standard layout is written for H = T. jar_1gal_180x197, at H/T 0.90, still
leaves 0.73 D over the upper impeller; jar_1p5L_112x215 holds 0.5 vvm only above 399 rpm, which
is echoed as a warning.
Not yet reconciled
Section titled “Not yet reconciled”- The baffle widths: four at T/10 against the T/12 the model sizes to; the sources accept both.
- Rutherford et al. 1996 is read through Li et al. 2007 and its abstract, not in full; the thresholds between its flow patterns are not claimed here.
- Rushton, Costich & Everett 1950 is read through McCabe, Smith & Harriott 1993, not in itself.
- The Handbook of Industrial Mixing’s pages are relayed, not checked against the text.
- The flow number: McCabe 1993 gives 1.3 for a standard flat-blade turbine (p. 245) where the model carries Grenville’s typical 0.65. The circulation comparison with a pitched blade moves with it, from about 1.8x in the pitched blade’s favour to near even (Agitation).
