pearl millet thresher - Recently Updated Pages

Work in Ghana

rasnow — Sun, 13 Dec 2009 11:59:30 CST

Dr. George Yaw Obeng, of the Technology Consultancy Centre, Kwame Nkrumah Univ. of Science & Technology, Kumasi, Ghana, submitted the attached report on Dec 13, 2009. The main conclusions and recommendations arising out of the study are as follows:

1. The performance of the millet thresher was high and was affected by the moisture
content and the rim speed.
2. The optimum threshing efficiency of 89.18% was obtained using the average speed
between 200 rpm and 325 rpm at moisture content of about 13%.
3. The germination of the threshed seeds was high implying that the thresher did not
cause internal damage to the seeds.
4. The grains that were produced were free from stones and sand as compared to those
threshed in the rural areas.
5. Increase in the number of spikes in the threshing head from 27 to 36 in order to
increase the force of impact by the spikes on the material; thus increasing the threshing
efficiency.
6. There is the need to design an appropriate threshing chamber to facilitate the
collection of threshed grains.
7. There can be viable enterprises manufacturing and selling the millet thresher after it
has been continuously perfected and scaled-up as a marketable product.

Return to Millet Home

High Speed Video at CSUCI

rasnow — Sun, 13 Dec 2009 11:53:13 CST

Work at California State University Channel Islands

rasnow — Sun, 13 Dec 2009 11:49:37 CST

We hope to better understand the physics underlying the revolutionary millet thresher. Our prototype also explores other improvements to the machine. With our simplified support system and alternate threshing surface, we are optimistic that the thresher can serve dual use as a functional bicycle for transportation. Our work in California is summarized below. A poster (pdf file) is also attached below.

Rasnow's millet garden.
The CSUCI thresher.
Thresher physics: high speed video.

pearl millet thresher Home

rasnow — Sun, 13 Dec 2009 11:46:18 CST

2009 Updates: See Work at Hampshire College, California State University Channel Islands, Mali, and Ghana pages.

=================================

This is the IDDS 2008 pearl millet thresher home page. The purpose of this project is to design and prototype an affordable, efficent thresher for pearl millet. This wiki documents our paths and progress during 3 intense weeks working in MIT's D-Lab to achieve that goal. Other D-Lab projects are documented on www.appropedia.org/IDDS and you can also listen to a report on NPR and see a video on The Christian Science Monitor.

Background: Pearl millet is a common grain foodstuff, called "bajra" in Hindi, "mil" in French, and "mahangu" in Namibian. Here are some videos of how the millet is pounded to separate it from the stalk and florets, and to make flour:

Girl Pounding Millet
Fulani girls pounding millet ***** March 2006
Fulani Women pounding millet March 2006
Ben and Seth in Mali - kids pounding millet
Senegal Pounding Millet and making flour

Threshing machines have a long history, although their design principles have been largely static since the mid 1800's. They come in a variety of sizes, prices, and throughputs. Nevertheless, much of the millet threshing in rural Africa and India is still done by slow manual pounding that consumes up to 4 hours/day of women's time, and results in blisters and repetitive stress injuries. Threshing is followed by winnowing to clean the grain.

Our goal is to make a radically simple, reliable, and inexpensive millet threshing (and ideally winnowing) machine, appropriate for small (<~ 1 acre) farmers. Ideally the machine will be built from locally available materials and parts, and be servicable with locally available skills, tools, and materials. It should be inherently safe, as young children are common in rural areas and often curious about machines. We are also interested in the applicability of the design towards threshing other small grains such as amaranth.

Problem Statement
Stakeholder Diagram
Research
Abstraction
Idea Generation
Reversal
Gallery sketches
Prototypes
2nd Design Review
Design for Final Presentation
Final Presentation pdf (attached below)
Contact us

Collaborating NGO's, Agencies and Individuals:
Jeffrey P. Wilson USDA-ARS, pearl millet researcher of 20+ years, has kindly shared his extensive knowledge on pearl millet as it is farmed, processed and consumed in Sub-Saharan Africa. He has also supplied us with an ample supply of pearl millet stalks, and been an enthusiastic supporter of our efforts.

Rolfe Leary, Engineer and volunteer with Compatible Technologies International (CTI). Rolfe and CTI volunteers have also been working to develop a low-cost method for threshing pearl millet. Jeff Wilson visited CTI in April 2008 and prepared A Review of Processing Options Developed by Compatible Technology International (see CTI report 4-10-08.pdf at bottom of page and a video). Rolfe has generously shared his insights and experiences related to this project and has been an enthusiastic collaborator.

Work in Mali

rasnow — Sat, 21 Nov 2009 12:02:52 CST

Dr Francisco Sebastian Rodriguez Sanchez has constructed and tested a prototype pearl millet thresher (PMT) in Mali and submitted the attached reports.

high resolution photos

rasnow — Tue, 06 Oct 2009 01:21:43 CDT

Two high resolution photos are attached below.

Contact us

rasnow — Wed, 10 Jun 2009 22:22:39 CDT

The IDDS 2008 Pearl Millet Thresher team consists of:

Donna Cohn, Visiting Assistant Professor of Applied Design, Hampshire College

Thalia Konaris
Michelle Marincel
George Yaw Obeng
Brian Rasnow, lecturer at California State University Channel Islands,
Francisco Sebastian Rodriguez Sanchez, (engineer graduated from Imperial College working in Mali)

CSUCI

rasnow — Sat, 28 Feb 2009 11:09:20 CST

The CSUCI millet page is here

CSUCI thresher

rasnow — Mon, 15 Dec 2008 15:01:40 CST

Applied Physics students Sam Hewitt and Luis Contreras helped build a prototype bike thresher demonstrating two innovations in the bike support and the threshing wheel.

Our stand proved surprisingly stable (we didn't even lock the handlebars), offers easy height adjustment for ergonomics, requires no modification of the bike frame, and is simple and cheap. It is constructed from two ~30" lengths of 3/4" EMT conduit with a ~30 degree bend ~16" from an end (note, none of these dimensions were measured). Four hose clamps attach the EMT tubes to the bike frame.

Our wheel had an aluminum rim, which we found cracked when we tried to bend the sides flat enough to access the millet to the spoke heads. Instead we purchased a strip of construction wire used to cover wall corners so drywall mud will adhere ($4/10' at Home Depot, but just about any wire mesh should do). We bent the wire loosely around the rim, tying it periodically between the spokes, and pressed the top with conduit to create a flat to concave threshing surface.

(Note the remaining piece of wire mesh is on the ground in the background).

The result was a pretty ugly wheel (and there was much skepticism that this would work). Also note the mesh is not horizontal/vertical, but diagonal, yet it doesn't matter. The threshing process is quite robust. We have little millet to test, but it stripped grain completely without stripping florets, from a couple of old and very dry panicles. This prototype throws grain fairly widely and we made no attempt to collect the grain, but will use the prototype for high speed photography of the threshing process.

It's hard to gauge the performance of this thresher relative to that of spoke nuts or the sides of the spokes. I suspect its performance is more efficient than spoke nuts because the wire characteristics interface with the millet better, yet it throws grain less widely than threshing in the spokes themselves.

Note that a tire could remain on the wheel below the wire. So a user could conceivably ride the bike to the field carrying the support tubes, hose clamps, wire wheel cover, and a reversible fender (which we haven't made -- as it will interfere with photography), and convert their bike to a thresher relatively easily.

The result, without rotating the stalk (I'm conserving my few precious stalks), is excellent.

Photos

lmp06 — Mon, 08 Dec 2008 14:17:21 CST

Slightly modified bike frame with leg extensions	Sam preparing to file down the legs for welding	Close-up of the filing
Another close-up of the filing	The fan component	Both the fan and the bike frame
Top view of the oil barrel with the shelling wheel attached	Close-up bike wheel	General lay-out of the design

Double V Belt Pulley welded onto old bike chain Close-up of the completed welding Another close-up

Working in the shop Placing the wheel into the barrel

High Speed Video at Hampshire

donnacohn — Fri, 05 Dec 2008 19:28:54 CST

Steve Banzaert graciously assisted us in shooting some high speed video using the least fancy of the fancy cameras he has access to. I believe this particular camera filmed at 2000 frames per second. We shined some very bright lights on the pearl millet. Excerpts of these videos can be seen at http://web.mit.edu/sgtist/Public/millet/. I posted a full version of the best video on youtube.

Work at Hampshire College

donnacohn — Fri, 05 Dec 2008 19:04:48 CST

Eights students in Donna's Appropriate Design in the World class are working on further iterations of the IDDS 2008 pearl millet thresher. They will continue the idea of removing only the grain from the pearl millet panicle, exploring the possibility of a thresher designs that incorporate winnowing and/or can be used by multiple users. The students will coordinate their efforts but have been divided into two subgroups.

Multiple User Design

Winnowing

High Speed Video

Winnowing

Braegan — Thu, 04 Dec 2008 19:26:03 CST

Braegan, Emily, Achyut, Marissa will focus on coming up with a method to incorporate winnowing into the thresher.

Braegan and Achyut are beginning their research, brainstorming and preliminary construction on a fan to blow away florets from the grain.

Emily and Marissa are researching winnowing techniques for all kinds of grains and even beans and eventually seek to build a shallow basket with mechanization to shake and shift the basket to separate and jostle the millet in order for the fan to be most effective.

For the Friday People:
Most of the fan is done. Here's where it is:
The pulley was too small to attach to the wheel, so it has been bolted on to this wood. The metal things with the screws through them are welded to the pulley on both sides. I think the weld was done well enough that the band will still fit through the pulley, but if not it should be OK to scrape off some of the metal blobs on the top side. There is more welding on the bottom.

The blades and the wooden thing need to be attached like this. Unfortunately, I didn't get this part done. It should be possible to put some of those little metal hooks (near the screw bins) around the spokes. That is the easiest way I can think of to do it.

These two wooden blocks will probably make a good base. The fan could sit on a metal rod where the blue line is. We've got a threaded metal rod already, but it isn't the best size (slightly smaller than hole). It would be better to have a new rod, but the one we've got will probably do well enough.

This is how the rod goes through the non-wheel part of the fan. This is just a bearing in a piece of pipe. The pipe is riveted to the blades. The whole thing is slightly off centered. That wasn't planned, but on the bright side, it may add a little bit of gusting.

The bearing can move around a bit in the pipe. I don't think it will come out, if it does then you can just stick it back in. Once the rod is through the bearing it should be impossible for the bearing to come out.

This is how the rod goes through the whole thing (minus the pulley).
If it is operating properly, the blades, wheel, wood part, and pulley should be what spins. The rod stays stationary.

Also, Achyut said he may some in tonight after his classes are done to look at it, so it may have progressed a bit.
Good luck, sorry I can't make it (I've got to hand in some projects that determine whether or not I pass a class).

Braegan

Past:

Our first plan was simply to build a fan similar to a floor fan. This would blow air over the florets, lifting the lighter chafe and blowing it away. We tried a number of shapes for the blades, but none blew hard enough. Frustrated, we decided to turn the blades completely sideways. This blows the wind quite strongly, but it blows it from the sides of the fan. This actually is closer to the design of a traditional winnowing machine, although we didn't know it at the time. We've decided to go with this design.

Present:

Currently, we are building the sideways fan out of steel. It will be attached to an aluminum bike wheel. Because aluminum is difficult to weld, it will be attached with rivets.

Future:

We still need to see how the fan will interact with the whole system. In a classical winnowing machine the fan is covered. In our tests we could not get the covered fan to push very much wind. Currently, we do not plan on using a cover.

Rasnow's Millet Garden

rasnow — Sun, 23 Nov 2008 22:12:26 CST

A small patch of pearl millet was planted 15 August 2008 in a 4x12' raised bed in Rasnow's vegetable garden (google map). Growth was fast, and flowers attracted lots of honey bees.
Here's a picture taken 10 October 2008.

A month later the seeds are forming and more children stalks emerged. Picture taken 15 Nov 08.

Bird netting on a frame was added before the seeds matured -- otherwise I am confident there would be no harvest. Here's my amaranth:

Multiple User Design

henrye — Mon, 20 Oct 2008 05:34:56 CDT

Andrew, Henry, Sam and Lilly will focus on the multiple user concept.

Multiple User Design sketch (hopefully integrating ideas of our group so far):

Research

seb.rdz — Fri, 12 Sep 2008 07:28:15 CDT

This section is divided in:

Father Lazarus ground truth from Namibia.
Background Information
Economic Information
Laboratory research (described in the prototypes section)

Background Information
(24 July 2008)
Summary of Conversations with Aaron Wieler (in India at the moment, but was in Namibia 2006-2007)
Stalks of Mahangu are stored short term in huts that look like round baskets on legs.

For most of the time, the pearl millet is stored threshed, but not winnowed. A family might rent a car to drive over their year's harvest of mahangu which would thresh it, but not winnow it. This costs around $15 U.S. Alternatively, they might pound it by hand in a specially prepared hole. In both of these cases, the grain gets mixed with grit and dirt and still needs to be winnowed to separate the chafe from the edible grain.

The grain is pounded into flour once a week. Another option is to carry a 10 liters (?) bucket of grain to the local Shimini each month, where it costs about $1.30 U.S. to get the grain milled into flour using an automated machine.

External links with information about millet and threshing

Wheat thresher from Hampshire College
Aaron Wieler - Bicycle Ambulance in Namibia
Performance evaluation of an Okra Thresher
FAO report on threshing
Thomas Jefferson Agriculture report on pearl millet
Pearl millet report by CRIAASA-DC for Namibian Government
Pearl millet consumption in urban areas in Namibia

Pearl Millet processing options developed by Compatible Technology International
Millet fact sheet from Ontario Ministry of Agriculture

Home Previous Next Contact Us

Idea Generation

seb.rdz — Fri, 12 Sep 2008 07:27:37 CDT

19 July -- this first round of idea generation was done by brainstorming. Individually we wrote ideas on post-it notes, and after ~15 minutes shared and grouped our ideas. This outline is the result. Ben also suggested we try reversal. Other ideas from patents search are also discussed.

▼ Idea generation
• ❑ lighter stick
• ❑ trampoline thresher
• ❑ shaking belt
• ❑ auto grain bag filler and weighing mechanism to stop feeding
• ❑ beat panicle; baseball bat
• ❑ blower or fan to separate chaff and grain
• ❑ conveyer belt (or other mechanism) to move chaff and stalk-- beating system also conveys so there's a continuous process
• ❑ conical feeder, or circular motion with one panicle at at time. Gravity feed. cone going outward so it doesn't clog. (peanut sheller)
• ❑ chute that sways, pendulum
• ❑ V shape to align panicles
• ❑ series of rollers, intestine,
• ❑ misaligned tires
• ❑ centrifuge
• ❑ washing machine
• ❑ water to separate and wash
• ❑ dissovle fluff -- biochemical
• ❑ 2 rubbing plates
• ❑ vacuum or pressure (vacuum removes skin of fish and avocado)
• ❑ compressed air jet
• ❑ vortexed wind tunnel
• ❑ woven mat
• ❑ Flower sifter -- squeeze the handle turning blades
• ❑ hypergravity
• ❑ Note: water might damage grain?

▼ Process steps
• ❑ feeding in
• ❑ separating stalk
• ❑ separating grain from floret
• ❑ cleaning, winnowing
• ❑ collecting grain, bagging
• ❑ collecting byproduct
• ❑ measuring quality
• ❑ maintenance
• ❑
• ❑
CAN we remove grain from panicle/floret without separating panicle from floret?

▼ Energy sources
• ❑ Air Pressure
• ❑ Torque 'em. Needed to push grain away from floret- shearing motion?

▼ Testing Ideas
• ❑ Stick a penacle into an electric drill and see how the rotation affects it. Use a bucket for a hammer mill effect
• ❑ Density measurements
• ❑ Vacuum pump
• ❑ Grind/ beat some materials
• ❑ Hammer mill test

▼ Problem:
• ❑ Convert penacle to clean grain
• ❑ Remove seed from floret
• ❑ Remove floret and seed from penacle/rachi
• ❑ Affordable and efficient thresher for pearl millet
▼ Remove and isolate grain

▼ Too physically strenuous/ hazardous
• ❑ Callous
• ❑ Repetitive motion
• ❑ Violent motion
• ❑ Dust
• ❑ Dirty grain when cooked, dangerous to eat

▼ Too time consuming
• ❑ Takes 4 hours to prepare
• ❑ Multistage production
• ❑ Lengthy pounding process
• ❑ Not a continuous process

▼ Lack of consistent quality
• ❑ Separation of grain from chaff by hand
• ❑ Tampering with sacks (lack of quality leads to even less quality i.e. no motivation to maintain it)
• ❑ Rancid grain if broken
• ❑ For large quantities rolling on the gorund collects dirt
• ❑ Placing penacles on the ground e.g. when drying

▼ Multiple stage process
• ❑ Spoilage
• ❑ Labour intensive

▼ Gender specific technology
• ❑ Burden on women who have other tasks
• ❑ Men can/do not assist with the process
• ❑ Affects repair and maintenance capacity?

Home Previous Next Contact Us

Ideas from patents

seb.rdz — Fri, 12 Sep 2008 07:24:29 CDT

Some extra ideas for threshing

Grain thresher
Cheese grater
Cheese greter 2

Links

seb.rdz — Fri, 12 Sep 2008 07:03:19 CDT

MIT News (August 8, 2008)

Design for Final Presentation

rasnow — Tue, 09 Sep 2008 01:09:08 CDT

6 Aug 08: IDDS 2008 presentations (millet and other projects), also reported by The Christian Science Monitor. A 1.2MB pdf of our final poster presentation is attached to this page, and a 750KB pdf of the final oral presentation is on this wiki's home page.

5 Aug 08: The final prototype millet thresher, ready for demonstration at the MIT Media Lab. The rim is covered with a scavenged bike fender, cardboard on the side with a plexiglass window to view the threshing. A short section of spring steel screwed to the fender presses the panicle to spoke nuts. Flags of Africa complete the prototype.

3 Aug 08: The bike spokes efficiently thresh millet, but the seed tends to go everywhere.

Video of the first bike spoke thresher.
Video of the first bike spoke thresher from through the wheel.
Video of the bike spoke thresher panicle ejection mechanism.

Also since high impact speed is critical, it made sense to try the spoke nuts as threshers. If the seeds stay within the rim and don't fall into the spokes, they will tend to not scatter as much. To contact the spoke nuts, the rim walls were bent outward with vice grips.

Video of the first bike rim (spoke nut) thresher.

From a piece of tin we created a "spring" to press the millet panicle against the threshing rim. Attached by tape to an old fender, with additional tin taped to the sides to reduce scatter of seed.

We raised the bike frame a little, making a stand with local materials. The grain is collected in a ... (drum roll) ... grain sack, clipped to some fender mounts. With some tuning we believe we can capture the vast majority of grain.

Before and After photos of the same panicle.

Home Previous Next Contact Us