Disclaimer: The experiment shown in this blog post is intended for informational and educational purposes only. Regardless of the identical experimental conditions for all specimens, the results are not being published in a scientific journal and are not being peer reviewed. For scientific data, please consult peer-reviewed scientific journals.
We recently introduced an AODD pump designed to pump trichomes. CFD simulations predicted low shear and gentle handling of trichomes based on the design and materials selection. We tested various particles to simulate trichomes and were pleased with the results. After a lengthy community discussion, we decided to do a simple test using trichomes to validate our claims. It was challenging to identify test conditions that would yield useful data. Some test conditions were proposed by the GLG community and expanded by us.
We took 3 pumps to test the difference between a Patented Low-Shear Centrifugal pump, our Hash pump, and a "normal" AODD pump as a control. We expected the centrifugal pump to be the best since it has few places to damage hash, with the data to support low shear, and a patent to prove its claims. We used a 1.5” AODD pump of a common design to compare it to our hash pump. We expected the "standard" design to be the worst.
We took 20 gallons of water and 500 grams of hash and recirculated that for about an hour. Each pump received separate lots of fresh freeze-dried hash which was recently extracted from a mix of biomass. Water was kept cold to prevent hash damage or melting. Hash slurry was recirculated for an hour. This re-circulation simulated pumping 4,000 to 6,000 grams of hash.
Though these tests are far from scientific, and you will not find a research paper here, we think the findings are surprising and worth the time and effort.
Many other test conditions can yield different results and we encourage the community to validate these claims and expand test conditions. Some things we would like to test are the following:
- How does the use of a damper effect hash on an AODD pump?
- What effects do different AODD speeds have on hash?
- What effect does AODD displacement have on hash?
- Do other AODD designs perform as well?
- Do other cultivars behave similarly?
Discussion:
Centrifugal Pump:
Data suggests centrifugal pumps destroy trichomes if deployed incorrectly. Keep in mind this is a Low Shear pump proven to work in other industries. Should you use a “Chugger” or other pump we believe this issue would be exacerbated. Until proved otherwise, we now strongly believe a centrifugal pump has no place in a hash lab.
We asked the manufacturer to comment on the matter and replied:
"The low flow rate you ran this model at, is a very inefficient point on this model pump performance curve. Less than 20% efficiency, at higher flows that pump would have run at about 71% efficiency. As a result, a tremendous amount of waste heat and shearing occurred. If it would have been a smaller model, it would have delivered the flow rate you required at 70% efficiency, and you would have seen much less shearing and much less waste heat transferred to product."
Ball & Cage AODD:
We now believe these pump designs are good for hash as expected, though more testing is required to validate this claim. Temperature sensitivity is still an issue and displacement are important. Running these at a slow speed is critical. I determined not to publish the results as it would misrepresent the pump's functionality. We will redo these tests in the future and publish the results.
Sambo Creeck Hash Pump:
Our hard work paid off and the data proves it. Much of the loss attributed to this pump may be explained by residues on the feed tank and hash inside the pump due to lack of flushing. Further testing to replicate the results are important. Testing other cultivars and conditions are also advisable.
Lessons Learned:
Many lessons were learned with this test and changes are in development for version 2 of the hash pump.
- The most obvious issue we ran into was size. Having such a large frame proved to be space consuming. Tests showed such a large frame was not necessary. A smaller frame will be used for V2.
- Air consumption was high on such a large pump. V2 will have a smaller pneumatic exchanger and use half the compressed air.
- A damper would have been useful, thus these will be offered for future pumps.
- Displacement probably does not play such an important role as initially thought. A smaller frame can thus be used. We still believe larger displacement is important for safety and slower pump rate at an acceptable GPM.
A smaller frame allows us to make a pump at 44 GPM for about half the price of the original pump. Ultimately these lessons learned are passed on as savings in CapEx and OpEx for the end user without a sacrifice in quality or performance.
Below are the results:
Shear Test 5/16/2022
Tested by Green Mountain Hash Co
Tech: Justin Massie
Test Conditions:
3 Pumps were tested:
- Low Shear Centrifugal Pump
- SC Filtration 2” AODD Hash Pump
- ARO 1.5” AODD Aluminum Pump
Target Temperature: 37F
Water Quantity: 20 Gallons
Ice Quantity: 20 Gallons
Room Temperature: 70F
Initial Dry Hash Weight: 500 Grams
Biomass Material:
A blend of CBDa/CBGa separated heads derived from 2022 outdoor Vermont hemp were washed to extract trichomes. Hash was Lyophilized to ND moisture and stored to remain granular. Strains include sour Tangie CBG, Peach Mint Kush CBD, Stormy Daniels & Cherry Bounce. Variance in chemotype and cultivar to provide various trichome sizes, densities, and stabilities to test against the provided pumps.
Process:
The lyophilized and pre-frozen material was allowed to re-hydrate and equalize temperature with the provided vessel for 15 minutes before engaging all 3 of the subject pumps, individually. The material was pre-stirred with a stainless-steel paddle to evenly disperse the material throughout the water before engaging the pump. Each pump was run for 60 minutes recirculating hash water back to feed tack at a room temperature of 70F. There were no markable environmental differences noted between each test. Each batch initially received 20 gallons of ice. After 60 mins recirculating, a water sample from each run was captured after a 20-micron filtration and each run’s resulting material was frozen and stored until all 3 were completed. All 3 tests were lyophilized at the same time.
Sieve Analysis:
All washed hash was sorted by size pre and post processing. Across tests 1, 2 & 3, no distinctive difference in shear was found across select micron ranges. It is likely that the size of a trichome is not a factor. The chemical content of the gland is a more likely factor in its likelihood to shear due to differences in structural characteristics. Tests (#1) that portrayed high shear tendencies exhibited indications in every bag used to collect (25, 73, 160um).
Chemical Analysis:
Samples were taken from the re-circulation water used and tested for Terpenes and Cannadinoids. No detectable cannabinoids or terpenes were found in any of the samples. One would expect at least the centrifugal pump to test positive since losses were substantial. We suspect there is a sampling error as the denser cannabinoids settled and were not sampled correctly or representatively. Further testing and different sampling methods are necessary.
Test Number: #1
Type: Centrifugal Impeller
Manufacturer: Anonymity Requested
Model: Anonymity Requested
Initial Dry Hash Weight: 500 Grams
Total time recirculated: 61 minutes
Starting water temperature: 37F
Ending water temperature: 43F*
Total gallons used: 20 gallons
Flow Rate: 10 LPM
Final Dry Hash Weight: 352.5 grams
Hash Loss Percentage: 29.5%
Notes:
High shear indicators were noticed throughout the entire process. During the rinsing/collection process foam was removed from the material that had previously been rinsed clean. accumulation of shorn heads was noted along the edges of the vessel, within the piping, apertures, and impeller housing. Within the first 6 minutes of recirculation, the water temperature had risen 3 degrees F. Over the next 30 minutes the temperature rose to 43f, at which point ice was added. The test concluded at 43F with rinse water at 37F. Collection bags also contained high shear indicators including blinding & staining.
*As a compromise, Test #1 was given an additional 10 gallons of ice due to a sharp temperature increase that otherwise would have ended the experiment as the hash would have melted or more easily damaged at warmer temperatures.
Pump Outer Casing
Pump Impeller
Initial Dry Hash
Final Wet Hash
Test Number: #2
Type: AODD
Manufacturer: SC Filtration
Pump Model: 2” Hash Pump
Initial Dry Hash Weight: 500 Grams
Total time recirculated: 63 minutes
Starting water temperature: 35F
Ending water temperature: 38F
Flow Rate: 15LPM Estimated (Due to the pump oscillation it was impossible to accurately read flowrate. 50PSI input was used for the closest comparable flow rate to Test #1)
Final Dry Hash Weight: 498 grams out
Hash Loss Percentage: 0.4%
Notes:
Very few signs of shear in total, a very light smudging across the edges of the vessel. Magnitudes less than the same place in Test #1. A small amount of material was found within the apertures of the pump and likely accounted for the 2 grams of loss. The temperature of increase of 3 degrees F took place for the first 6 minutes of recirculation. The following 54 minutes showed no measurable increase in temperature. During the middle portion of the test, the pump seemed to exhibit high shear indication in the form of foam, However, this may have been the result of already shorn heads that had been collected in the original process.
AODD Hash Pump Setup
Iced Hash Water
AODD Outlet
AODD Inlet (Possible Cause of Partial Loss)
Starting Ice in the vessel.
Hash water was recirculated back to the tank. Notice foaming.
Hash water near the end of the cycle.
Residue on Tank. This was consistent in all tests, though more on some. It is a possible explanation for losses.
Hash to Lyophilization and final weight.
Test Number: #3
Type: AODD Ball and Seat
Manufacturer: Dayton
Pump Model: 6PY44B (Aluminum)
Initial Dry Hash Weight: 500 Grams
Total time recirculated:
Starting water temperature: 35F
Ending water temperature:
Flow Rate:
Final Dry Hash Weight:
Hash Loss Percentage:
Notes:
After some debate over the performance of this pump we decided the results were inconclusive and the experiment should be repeated.
Dayton AODD Pump Setup