High Pressure Hot Air Blower - Sound Containment.

NXT Roasters started with off-the-shelf centrifugal blowers of which we got to know the internals too intimately.  There were no guarantees that they would handle the temperature but the salesperson thought that we should get a reasonable MTBF (Mean Time Before Failure).  At the beginning, it was only 25 hours on average!

Virtually every manufacturer claimed that their composite bearings would work at the 250°C air temperature.  They must not have realized that 18,000 rpm was on the faster side which greatly reduced the life of their product.  Regular bearing races are too soft for this temperature and speed combination.  It takes very little Brinelling which creates a looseness that, at 18K rpm, allows  the shaft to vibrate and the bearings self destruct.  Note: the faster the speed the smaller the movement of the vibration, all other things being equal.  Bearings were repacked with high temperature grease that was measured out in milligrams.  Oil bathed bearings were tried.  Many iterations of new seals had to be turned because the stock items were intended for ambient temperatures. Most modifications succeeded in improving the longevity; 25 hrs, 55 hrs, 125 hrs, 225 hrs, 320 hrs, and after a couple years it was over 1300 hrs.  Not that great, but the client could live with it while we were working on a new blower that is featured in the other blog pages.  A specialized bearing manufacturer made a bearing for us that goes like the proverbial bunny.  All we have to do is pay them a lot of money.  The switch over to our blower design started two years ago.  We do not know the MTBF because of the lack of failures (no! no! that is good).

The aim of this blog is to discuss the blower noise that was generated by that first blower compared to what we have today.

The off the shelf blower had an aluminum Belt Guard with a rubber lip that is not shown.  The guard resonated and the RoastMaster wore Ear Protectors.  Obviously the level of balance is greatly affected by the speed of the 6" Rotor.  If that manufacturer recognizes his cover they will realize that they have a problem and then see the solution.  Well, they have more problems than that which is why they do well to focus on ambient temperatures.  The paint has flaked which reflects badly on the quality of the work.

The output of our design was better than expected therefore we were able to reduce the speed.  To improve upon the case design, first, we increased the mass by making the Belt Guard from steel sheet.  It was better, or as they say, "Good, but not so good".










Heavier bar stock was screwed to the front and sides to increase the mass and reduce the amplitude of the resonance.  The roaster was delivered with that modification but the client still found the noise to be disturbing to the working environment.

In his youth the client was a hot rodder and had some experience with taming unwanted frequencies though I suspect that most of his energy went into making the engine louder.

He covered the Roastaire Belt Guard with an aluminized Butyl Rubber sheet product.  It worked so well that the covering panels got the same treatment.  The "sound treatment" was so effective that CBC Radio did part of "On the Island"  Roastery Interview in front of the working Roastaire.

At the Eptech show Ampco's Rick Gagne discussed such a product that they carry.  It is not necessary to cover much more than 25% of the surface to stop the ringing but esthetics would take a beating with less than full coverage.  The alternative was to place strips on the interior walls but if the bond fails after a while then that will certainly lead to belt failure.   Perhaps we can revisit and find a conclusion after a year - with the help of a freshly roasted coffee.

Vented Bags Revisited

The original blog of  Do Coffee Bags "Lock in the Flavour" ?   and Packaging Optimism gathered the most visits.  Does this mean that those people believe or want to believe that there is a magical way to preserve roasted coffee.  Well, apparently there is but it involves a -273°C or -459°F refrigerator.  That is no ordinary ice box.  Even then the degrading chemical reaction is only slowed down not stopped.

This little valve is, in my view, good engineering. The object can be mass produced cheaply and there is much more to the design than the average coffee drinker pays attention to. Notice the fiber filter which is backed by a plastic frame to give it strength. I assume that the purpose is to keep coffee particles away from the sealing lip but then it may also keep small life forms out of the bag.

Interestingly, the valve is required to preserve the bag rather than the coffee.  In the days after roasting the coffee beans will produce enough CO2 to burst the bag ( it depends on the bag material but I found that 250 to 350 grams was the limit for un-vented vacuum packed bags).  I suspect that the "longevity improving" aspect was more an invention of the marketing department than an engineering criteria. 

A marginal improvement in longevity might be possible with vacuum packed vented bags.  If the bags are packed within the hour after roasting they will need the valve even if they are vacuum packed.  The porosity of the bean is relatively low and it takes time for the CO2  to migrate from the bean.  Without a vent even vacuum packed packages look like overstuffed pillows.  Obviously that would not allow efficient boxing of the bags.  

Rather than believe that the bag was not kept in a warehouse for months and or spending days in truck trailers I prefer to buy from a local roastery or from Fresh Cup because they expedite the order on the day it is roasted.  Oh well, that offer is only for Canada which means that more than Alaska and Hawaii are hit by the shipping restriction.

The Importance of Timely Bean Cooling

Coffee is roasted by way of an exothermic reaction.  Without cooling, the beans would invariably burn because they reach the combustion point.   Roasting consistency can only be achieved through the timely introduction of the cooling air, which halts the roasting, to all the beans.

Drum roasting requires well developed skills because there is an extended lag between the moment the RoastMaster views the beans in the Bean Sampler and the termination of the exothermic reaction.  Several factors affect the degree of roast in a drum roaster; the air temperature is considerably higher than what is required for roasting  ("Espresso Coffee" R. Eggers writes the typical gas temperature of a drum roaster is 400° - 550°C. ), using a Bean Sampler is a relatively slow procedure, and the Cooling fans are relatively inefficient at moving air through several layers of beans.  Sucking the air downwards through the beans is also bad design because it packs beans together which greatly hinders air flow.  Many drum roasters spray water on the beans to cool them more quickly.  This is not good nor is it a measured cooling method.

The high temperature cooking analogy for beans hit by the super hot air is cooking Scrambled Eggs while using the stoves highest heat setting.   Even if they are continually mixed the heat is much higher than the egg can transmit to the interior.  The higher heat of the drum roaster still yields a relatively slow roast which highlights the inefficient transfer of energy.         For more details see "NOT Sustainable Roasting" and "Heating Coffee Beans to Roasting Temperature"

It is said that every second counts when "calling the roast" yet it takes many seconds to determine the degree of roast using a Bean Sampler especially if multiple readings have to be taken to follow the progression.

Many roasters are  stressed  by "calling the roast" even after they have been roasting for nine months.  The beans continue to roast during the dumping process and at the beginning of the air cooling cycle.  It is a common drum roasting technique that water is sprayed unto the beans to soak up heat.  It is impossible to accurately meter water coming from a garden hose.  Moisture is the #1 enemy of roasted coffee.  That lingering roast period means the drum RoastMaster must anticipate the roast quite a bit in advance of the required roast level.  An appropriate analogy is shooting ducks; the expert hunter aims for the position where the duck will be when the buckshot arrives.  The hunter must compensate for altitude, shell powder load, temperature, duck speed, and probably many other things that only a good hunter would know. For various reasons a lot of people have a problem with duck hunting (no ducks were harmed for this analogy).  The drum RoastMaster does not deal with less variables.

 To this end the NXT aka Roastaire uses a fluidized bed in which 240°C (adjustable as required) air is pushed, under pressure, through the beans.  This significantly increases the heat transferred to the entire batch of beans.  The beans have a limited ability to absorb the heat therefore the temperature gradient is more uniform throughout the batch with the forced air stream.  The added benefit is that the air flow continually churns the beans and carries away the chaff.

 The beans are cooled within the Roasting Chamber therefore "calling the roast" is done in real time.  The blast of cooling air also churns the beans so that there is even and rapid cooling.  The Roast Monitor communicates to the RoastMaster the roasting progress so that the Cooling Cycle is started at the proper moment.  It only takes a few seconds to cool the beans ( below 180°C according to R.Eggers and below 200°C according to Clarke & Macrae) to stop the roasting process.  After this 30 second Cool Mode the beans are transported to the Packaging Hopper while the NXT aka Roastaire loads nother batch freeing the RoastMaster to appreciate a freshly roasted coffee.

Carbon reduction or is it just "horse trading"*

There is even more to the story !

The unfortunate part about marketing claims vs. true development and technical improvements is that the marketing claims focus on changing perception rather improving the equipment.  It is the easy way out which in this case does nothing for the local air quality if someone planted trees hundreds of kilometres away.  The article cited does not mention that it took a decade to reduce the Roastaire™ emissions to a small fraction of conventional roasters.  In the process the energy required to roast was reduced to about 1/6 of a conventional coffee roaster.

The Roastaire™ was designed for the Café environment where people can observe the roasting process.  We refer to that as the "theatre of roasting".  The roaster is a compact unit that includes Chaff Cyclone, SmokEater™ (pollution controls), Pneumatic Bean Conveyor, and a Packaging Silo.  The roaster is approximately 6.5 feet high.  The roaster is controlled by a computer based Roast Monitor which  also configures the roaster, through the use of pneumatic valves, for the various modes in the roasting process.

More information is given in the post "Sustainable Design Follows Function"  such as the attention given to ergonomics.  An obvious example is the placement of the Loading Hopper which is placed at counter height rather than above the RoastMasters head.  The roasted beans are quickly cooled internally so that the smoke can be neutralized before it is vented.  The aim has been to improve and streamline the roasting session.  Cooling with a sustained blast of cool air within the Roasting Chamber means that the RoastMaster does not need to compensate for the roasting that takes place between the time of the decision and the moment that the beans are lowered below their combustion point.  The sum of the drum RoastMaster's reaction time, Roasting Chamber evacuation time, plus the actual time it takes to sufficiently cool the beans to stop the roasting process means that the RoastMaster must anticipate the final roast a minute or more in advance.  In drum roasting the process is analogous to "duck hunting";  if the hunter aims for the duck they will miss.  The hunter must shoot ahead of the duck by a distance that varies with variables such as height and speed of the duck.   Similarly skill and much training is required to anticipate the time required to get the correct drum roast.  The drum concept does not make it easier for the RoastMaster but that analysis will have to wait.

To help improve the efficiency of the Roastery the roasted beans are pneumatically conveyed to the packaging area where they can be weighed and placed directly in the bins. 

The Packaging Hopper can be mounted to the wall or placed on a table or counter.  It can be placed directly over the scale if the RoastMaster finds it more convenient.

Café owners that buy roasted beans see " FRESH Coffee! " post are simply paying for someone else's roaster without getting all the benefits.




* Horse trading is an idiom used to describe negotiations, especially where these are difficult and involve a lot of compromise.

High Temperature Pressure Gaskets

Several factors have increased the sealing challenges:
- temperature of 250ºC
- a system pressure of 1.1 psi when at roasting temperature
- Hot rolled flanges, not machined surfaces
- it must provide a barrier to very acrid smoke
- must resist suction
- must not out gas after burn in
 

Silicone sheet was touted to withstand the temperatures but over the years it crystallized and became brittle.  Some Teflons have a fairly high temperature ceiling but most are not nearly as capable and durable as the claims make them out to be.  We attempted to use high temperature ceramic paper but by itself it turned out to be too porous.



Out of curiosity we added a ring of packing material which would further compress the ceramic fibers but it still leaked.

It was possible to seal machined faces with High Temperature paste, some of which was "scaringly" smelly but most of the flanges are cut from hot rolled sheet material.  Better sealing capacity was required over a slightly wavy surface which is exacerbated by the weld to the tubing. 

Samples were cut from various materials that were either on the shelf or purchased.  The initial tests consisted of about four times as many gaskets as shown in the picture above.  A rather simple device was salvaged from the "steel pile" to pressurize the gaskets.  An arbitrary 1.5 atmospheres was selected as the testing pressure.  A soap mixture (sorry I forget the soap's brand-name  which pretty well ruins an endorsement contract) was used to dramatize the results for the camera.  The relative size of the bubbles is proportional to the size of the leak.

We looked for sealing ability especially after being subjected to repeated hot/cool cycles but also the material had to resist extrusion.

Many of the gaskets sealed until they were subjected to periods at 325°C a temperature that was also arbitrarily chosen to represent heat damage over an extended period.  Some people might complain that this is not a fair test.  It might not be but since their product is not named they will not know that they should complain about the procedures.

To the left is a Silicone product that is resistant to 350°C** (you have to search the fine print for the limiting conditions of only being flashed to the temperature on a Wednesday morning between 9:03 and 9:08 AM".  The extruding material created an even greater problem and it was discarded.



To the right is the outside view of the testing oven.  A PID Autonics temperature controller was added with a RTD sensor.  As mentioned above the Set Point is 325°C.  This little throw together may not win an award (it is from the same design studio as the testing jig to the left) but it keeps a stable temperature. 

Seal progress was made through the combination of various products but further work is required to determine the mixture and the method of application so that a uniform gasket is produced.

Personally - I prefer to drink freshly roasted coffee!

High Temperature Blower - Streamlined Inlet

This is another picture of the back of the High Temperature Blower which better shows the Inlet Flange.  In this version the Inlet is attached from inside of the case.   We are currently fabricating an updated version which will be fastened from the outside.

We are working on moulded ceramic insulation but those efforts are not yet worthy of a picture.  To date we have used Cerawool Blanket, ceramic hardboard, and aluminized high temperature fiberglass.  The process is tedious because of the numerous cuts (not all on the insulation) and the challenge to fitting the material to such a "curvy" object.

The unit pictured below has been working 8 hours at 250°C  nearly everyday.  The version, not using the Curved Inlet, is nearly two years old but is used more lightly, though at the same temperature.  A third unit has been used for testing purposes but will soon be used on a nearly daily basis.

A new Inlet Flex Connector has been fabricated and it will soon be tested.  The aim is keep the sound level as low as possible.  We have had a measure of success in accomplishing that.  Last week CBC Radio aired an interview that was done just in front of the working Roastaire.   Obviously more improvements will be implemented to make the work area as pleasant as possible.