my forge-development thread

[jwmelvin]

I had a go at forming a refractory lining for the forge from KOL30. It was pretty dry and difficult to form into a uniform shell. I used poster board inside to support the inner surface, and stuffed that with crinkled paper. Around the openings I applied blue tape, which had a hard time sticking to the damp poster board, but helped quite a bit to get the full opening supported. I added a little refractory to the nozzle I had earlier cut out and cast separately, and then with paper between the nozzle and forge-body refractory, inserted the nozzle unit into its opening. My plan is that I can later remove the nozzle unit and replace it with the ribbon burner. The whole assembly is now wrapped in a garbage bag to set.

[Frosty]

If you do this again or make mods add  more water! The % moisture recommended on the site are for max strength and heat, those far exceed what you need in a propane forge. Don't get silly with the water but unless you have a mechanical mixer it's almost impossible to work it with the "recommended" moisture content. 

Were I making forge liners very often I'd modify a muller or make an auger mixer. By hand it's a PITA, you need a BIG mortar and pestle more than a spoon.

I'll be looking forward to seeing this baby up and HOT. 

Frosty The Lucky.

[jwmelvin]

Thanks Frosty. It did seem easier to work with more water (my first batch) but then I got concerned I would ruin everything by overdoing it so went with 18% hydration. I am still hopeful that it will turn out okay. I will try to fire it up in the next few days but have a busy weekend ahead with the family. 

[Frosty]

Yeah, I went that route myself but after making my arthritic thumbs hurt I decided more water!  Even with enough water you start getting free moisture it's hard to mix thoroughly. That is in large part because the aggregate is all crushed so the sharp angles key together and don't slide past each other easily.  An exaggerated example would be two equal sized bags, one filled with marbles the other filled with jacks. One shifts easily the other not at all. Jacks key together. Make sense?

Do you have it sealed in a plastic bag with a WET towel so it cures properly? It really needs 100% humidity to cure properly. Over night is good but a week is recommended max effect.

I like linoleum for forms but you need to grease it up or Kastolite sticks to it HARD. 

Frosty The Lucky.

[jwmelvin]

Yes it’s in a plastic garbage bag so should be pretty high humidity. The contents were pretty wet. Linoleum is a good idea. I am hoping I can burn my poster board away so I’m not too worried about it sticking. I’m excited to take the next step but it sounds like a day or two of sitting around will only help it. Regardless, I know more for next time. I hear you about the nature of KOL. That’s why I’m glad I have another material to try for my ribbon-burner block. (I’m printing a new mold for that right now.)

[Frosty]

Just because the Kastolite was still wet doesn't mean it doesn't need more water. It does NOT dry, the binder and water for chemical bonds after it sets to cure to full strength. Portland cement concrete continues to get stronger the longer it's in contact with water. some of the strongest concrete on earth was poured by the Romans the Greeks didn't quite have the recipe right.

Kastolite behaves just like Portland cement concrete so toss a wet towel in with it.

Frosty The Lucky.

[jwmelvin]

Over the past couple days I have added KOL patches to some areas I didn't like in the initial attempt. KOL, once set overnight, is pretty crumbly. I was careful as I went over it, but a couple sections were too thin and came off. I mixed a couple batches of patch, partially because I wanted to know how well it would adhere to the set-but-unfired refractory. For the second patch batch, I mixed the KOL quite on the wet side. Today I baked the forge in the oven, ramping up to 500 ºF over five hours. 

I tried to fire it with the refractory nozzle I cast, but it was too little expansion and the flame wouldn't stay lit. I had read that one doesn't really need a flame-retention nozzle once in the forge but I couldn't get the flame to stay lit. So I shaped an IFB to fit the entry port and drilled a hole in it for the 1.5" stainless-pipe nozzle I have been using in free air. That worked well:

I adjusted the nozzle so the flame front was around the mouth of the nozzle and let the forge heat up for a while at about 5 psi. It seems good to me:

Sometimes the fame front detached from the nozzle but it never went out:

After a holding temp for a while, I increased to 10 psi and it got a lot hotter:

Then I turned the burner off and the color of the KOL lining seems to indicate that the temperature distribution is reasonable:

I put the ribbon-burner casting in on a tray to let the PLA core burn out, which it did nicely running as low a pressure as I could:

So now I have to finish making the plenum for the ribbon burner, then I can test that. Next time I will try to hook up the thermocouple too, as it seemed to work well to cast a port through the refractory for the thermocouple sheath. Pretty soon I will coat the inside with Plistix 900F. I think there is no reason to wait for that now that I've fired the KOL. 

[jwmelvin]

I fired up the forge for my first attempt at bashing metal tonight. It seemed to work pretty well. I was running it at about 7-8 psi. I was too excited to take many pictures. Here is the flame on startup:

Once it got really hot it had a hard time pulling air. I think the mixing tube was too short and the inducer softened so that it slipped out of alignment with the mixing tube. It worked fine after I let it cool, lengthened the mixing tube, and continued. The mixing tube felt cool after I restarted. Perhaps the forge temp was high enough during the first burn that it changed the combustion somehow? I was trying to open up the ends to make sure I wasn’t choking it but it had already impaired the inducer at that point. 

It was fun trying to make a pair of tongs:

[Another FrankenBurner]

Your nozzle ends at the forge cavity which will cause it to over heat.  If you can pull the nozzle back into the firebrick a ways, it will help.  When I was using metal nozzles, I liked the outlet of the nozzle to be outside of or flush with the forge shell.  Recessed beyond the refractory and the blanket.  

Another thing you might have going on is a nozzle which is too large.  That image looks like the flame is back in the nozzle which also over heats it.  With a metal nozzle, you want the diameter small enough that the flame rides at the outlet of the nozzle but big enough that it prevents flame lifting.  

This over heated nozzle then conducts heat up the mix tube.  

Nice looking tongs for a first time.  Is this your first forged project of any kind?  The jagged markings in your offsets are because the edges of your anvil are so sharp.  

[jwmelvin]

7 hours ago, Another FrankenBurner said:

 If you can pull the nozzle back into the firebrick a ways, it will help.  ...

Another thing you might have going on is a nozzle which is too large.

Thank you I will certainly try both. I could see the nozzle was too hot and pulled it back later but not as much as you suggest. 

Yes first time beating hot metal. It’s not easy!

[Mikey98118]

Um, BTW...congratulations on a very nice flame in a very nice forge; we don't want to forget mentioning that in our rush to help you

We didn't use to have an embarrasment of reaches, but this year so many of you are doing a grand job with your first antempt that we may start takcking you for granted 

[jwmelvin]

17 minutes ago, Mikey98118 said:

congratulations on a very nice flame in a very nice forge

Thank you Mikey. I am happy that it's working so far, though it has plenty of ground for improvement. Though it's all so new to me that I'll need to decide how to split my time between learning to work hot metal and improving my tools. I suppose that's pretty common with many such endeavors...

[Another FrankenBurner]

Thank you for the reminder Mike.  I get ahead of myself and forget my manners from time to time.  

It is a beautiful blue flame and a hot looking forge.  Nice job.  

One thing I just thought of, if you pull the nozzle back into the fire brick, the firebrick may be heated more which could cause the brick to crack. 

 

[Mikey98118]

A thing of beauty may not be a joy forever, but we must take what  we can get

[jwmelvin]

1 hour ago, Another FrankenBurner said:

One thing I just thought of, if you pull the nozzle back into the fire brick, the firebrick may be heated more which could cause the brick to crack. 

Yup, that’s what happened today. I figured it would be an issue and decided I should line the entry port with KOL. That was my original intent but I made the nozzle too small and it won’t stay lit so I switched to the SS nozzle through a firebrick. 

At various times with pressures below ~5 psi, I get some oscillation in the induction. Not the crazy chuffing that I’ve seen with other conditions but still made me wonder. I will video next time. 

[Mikey98118]

All naturally aspirated burners have minimum and maximum gas pressures for adeqate flame control; we call that their turn-down ranges. Huffing is a strong indication it's time to turn up the gas pressure a little.

If you can't turn down the gas as much as desired, it becomes time to build a smaller burner; we all do it

[Another FrankenBurner]

The cast nozzle is the way to go.  I have completely abandoned metal nozzles.  I recommend casting a few in cheapo plaster to find a good size/shape so you don't waste KOL.  I prefer to cast the nozzle on it's own and then cast the forge liner to glue the nozzle in place.  Printing the form and trying to ram KOL in place in the forge hasn't turned out as well for me.  

[jwmelvin]

Last night I had a go at making an expander plenum for my ribbon nozzle. This was to please my aerodynamicist brother, who thought the sudden expansion from a mixing tube to the plenum was a waste of energy in the flow. I'm not much of a sheet-metal fabricator but it was fun to play around, and I came up with this:

It would not run in free air, as the outlet velocity was too high (video here). So I tried it in the forge and it was stable, though had a fairly uneven pressure distribution (you can see that the flames at the bottom of the ribbon were being blown much farther from the block):

I could see that the orientation was not ideal (it was pointing too much at the adjacent forge wall) so I reoriented it and it looked much better (all this so far is at about 5 psi):

Once the forge warmed up, the flames looked better (to me). I ran at a series of pressures and it seemed stable from about as low as my regulator could go up to the 10 psi I tried:

The forge was still not that hot when I did this, so I'm not sure how it will perform at true operating temperatures, but it was exciting. I'm sure some of the flame nonuniformity comes from me melting the core a bit when prepping it for casting; I was using a heat gun to try and remove the strings but that was ill considered; I tried to straighten them as much as possible but they were nothing like the perfection of the original print. I'm sure I have some nonuniformity in the pressure at the back of the block, as the expansion is too severe, but I wanted to constrain the overall length. I haven't yet tried different mixing tubes and I know that the inducer-jet assembly is not all that well aligned. So some further tuning should help. 

Here are some videos of it: (1) running on startup; (2) running on startup once reoriented; and (3) running once warmed up a bit, each at about 5 psi. 

I welcome any guidance or comments. Especially on what to look for when tuning the burner in one direction or another. I need to add a needle valve to better regulate the low output. 

[Frosty]

I don't think aerodynamics and saving energy is a good concern. You need turbulence to distribute fuel air evenly over the outlets. My flames didn't burn evenly till I put the inducer inlet at 90* to the outlets, there's nothing aerodynamic about slamming the flow against a steel wall 2" from the inlet.  That said, it's only an opinion and I'm wrong so often I've stopped being surprised about it.

I think the flames are running a little rich which makes me wonder at your inducer build. You might want to try a Side arm or T inducer, they produce better outlet pressure which SHOULD(cross fingers) make a stronger flow and let you adjust psi across a wider range. They're also much easier to adjust air fuel ratio.

It's looking good so far though.

Oh and rather than announcing you're going to update your development thread in other threads please just post the update. Folks interested in ribbon burners are following your progress but if we keep opening posts just to see a link to another post we might lose interest in opening any of them. Make sense?

Frosty The Lucky.

[Buzzkill]

Take this for what it's worth, but in my experience there is a benefit to flame stability by introducing the mixture to the plenum at 90 degrees from the burner block.  There still always seems to be spots where the flame is higher from some holes than those nearby, but in general it seems to produce better results than aiming the mixing tube directly at the burner block.

The other thing that may make a difference is the number and/or diameter of your nozzles relative to your mixing tube/inducer setup. Since your setup appears to be able to pull in more air and/or mix it better than some other options you might be able to run your ribbon burner with a smaller diameter tube and maybe even a smaller mig tip.

Do you feel like the pictures you posted capture the colors of the flames you were seeing accurately at the different pressures?  I generally try to shoot for flames about the color of your 5-6 psi picture.

[jwmelvin]

Thanks for the thoughts. The poor result with introducing the flow to the plenum along the axis of the outlets is exactly why I made the diffuser. One would not expect a good result (i.e., good pressure distribution at the back of the ribbon block) from just a sudden expansion <2" from the block. My diffuser is steeper (more severe of an expansion) than ideal for expanding flow (~6° included angle) but it is in the best shape to prevent separation of the flow from the walls. I think there is not much back pressure from my ribbon block, which would make it easier to expand the flow and achieve a uniform distribution. 

I hear what you are both saying about a 90° turn improving uniformity, but that will reduce the overall flow, which I am trying to avoid. It is a compact and simple way to achieve the same thing and I may end up there eventually. Just playing with some other ideas for now. 

I may try feeding the nozzle with a smaller burner (which I agree would require a smaller gas jet). But I guess I'm wondering what I would be seeking through this: better efficiency, or a lower bottom end? I don't yet know what steady-state temperature I will get from running this at its lowest output. 

I will pay more attention to the color of the flames in the pictures as compared with what I see. As I remember it, they are reasonably accurate. I remember thinking the low pressure looked quite rich, as reflected in the greenish hue of the flames. The mid pressure looked reasonably good to me, as you say. I figured that would be a typical operating point for the forge so that seemed good. The high pressure was just kind of unstable and had a lot of dragon's breath. I do have some tuning to do on the inducer setup. 

My intent here is to determine whether one really requires turbulence to achieve a uniform pressure distribution. While that is one way to do so, it is not the most efficient. As you suggest, however, it is a way to achieve mixing and homogenize the pressure distribution. I'm just trying to work on another approach. I may try some tests to see the difference though.

Why does a side arm or T inducer produce better outlet pressure? Better than Mikey's style inducer or a linear inducer? I haven't begun to tune my inducer setup for this nozzle, so I'm not ready to jump ship quite yet. 

As far as updating posts, my intent was to post content relevant to the other threads. I can stay out of your NARB thread if you prefer; that was an update and pointer, and I can see how you prefer I not do that. The flared ribbon-outlet thread was about a particular topic and I think my post there was both relevant and responsive to a prior post. I intentionally did not post media in those updates because I know that you are sensitive to such things. I didn't realize quite how sensitive.  I do appreciate your feedback here though, and hope you don't find it too burdensome. 

[Frosty]

I'm following your progress, my suggestion is only to let you know it's not necessary to announce your progress reports. 

I thought you were looking for critique evidently I was mistaken. My apologies, I'll just read and look at the pics. 

 Frosty The Lucky.

[jwmelvin]

Oh, I am definitely looking for critique. And I do appreciate your input that a 90° turn is effective; like I said, I would like to try that to see the difference. My hope was to motivate a few people to think about new ideas. 

[Mikey98118]

What I seem to be seeing is sharp clean burner flames. You may work on the icing a little more--or not--but you've plenty of cake

[jwmelvin]

Thanks Mikey. It does seem plenty usable and ready for some temperature testing; I just enjoy playing with the design to figure it out a little better. 

I did a few experiments for better mixing by using metal window screen in the diffuser, placing screen: (1) flat across the end of the diffuser; (2) same, but with +3 layers on the outer ~1.5" of width; (3) flat at the start of the diffuser expansion; and (4) rolled up in the end of the mixing tube. Here are pictures of (1), (3), and (4):

Here are the flames that resulted:

It was amusing how much the rolled screen choked the flow; I thought the secondary-flame vortex was neat so I took a picture. The flat screen in the diffuser didn't seem to affect flow much at all. It also didn't do much for the pressure distribution. But adding screen at the sides of the diffuser did help tame the pressure there, to the extent that I think it's an improvement. I wonder what extending the mixing tube into the diffuser would do; I imagine it would move back toward a center-heavy distribution...