Japanese Shells

[wizard7611]

How is this shell made(0:37)?

 

https://www.youtube.com/watch?v=SRVYagcmQnQ

 

Oh, I almost forgot. How do you roll ghost shell stars?

 

-Thanks!

[pyrokid]

Those shells are all 12" diameter so they have some room to work with inside the shell. It has become quite popular recently for the Japanese to make use of empty space in their shells. Often, instead of a full peony, they will use a more sparse effect that is made with less stars inside the shell.

 

See :https://www.youtube.com/watch?v=dA8dnFuY2y4 Imagine how you might create this sparse effect by starting with a normal peony and then removing some stars. I think generally smaller stars are used to create denser groups than what are used in normal peony shells.

 

In the case of the snowflake, you can see a few different petals:

 

1) Large white comets; outermost

 

2) Color changing stars that follow the comets most closely

 

3) multicolored groups that follow the color changers; seen clearly at 0:44

 

4) White inner petal

 

It might be productive to view the stars in terms of stacks of effects that extend radially from the center.

 

 

Ghost stars are color changing stars that are rolled precisely. Different batches are made with different timings, and these are assembled into the shell.

 

The Japanese are masters of the craft; it is their mastery of each of the steps in shell production that makes these effects possible, to say nothing of the artistic vision required.

 

Also a note: here's the 2015 competition! These are always so exciting, I always see an effect I've never seen before!

 

Edited November 29, 2015 by pyrokid

[dave321]

^{as you say, the Japanese are masters at shell construction}

^{some beautiful effects there}

[wizard7611]

Those shells are all 12" diameter so they have some room to work with inside the shell. It has become quite popular recently for the Japanese to make use of empty space in their shells. Often, instead of a full peony, they will use a more sparse effect that is made with less stars inside the shell.

 

See :https://www.youtube.com/watch?v=dA8dnFuY2y4 Imagine how you might create this sparse effect by starting with a normal peony and then removing some stars. I think generally smaller stars are used to create denser groups than what are used in normal peony shells.

 

In the case of the snowflake, you can see a few different petals:

 

1) Large white comets; outermost

 

2) Color changing stars that follow the comets most closely

 

3) multicolored groups that follow the color changers; seen clearly at 0:44

 

4) White inner petal

 

It might be productive to view the stars in terms of stacks of effects that extend radially from the center.

 

 

Ghost stars are color changing stars that are rolled precisely. Different batches are made with different timings, and these are assembled into the shell.

 

The Japanese are masters of the craft; it is their mastery of each of the steps in shell production that makes these effects possible, to say nothing of the artistic vision required.

 

Also a note: here's the 2015 competition! These are always so exciting, I always see an effect I've never seen before!

 

Thanks! So, how do you remove peony stars without them falling and grouping together again? Also, for the ghost effect, are color changing delay compositions used to create this effect?

[pyrokid]

The removal of stars from a peony pattern was a conceptual thing. There are a few different ways. I'm not sure what the most popular method of shell assembly is in Japan. There is the two hemisphere route, there is a variation of this, where more than two "hemispheres" (really less than a full hemisphere) are assembled to form the sphere, and there is the method where the shell is built from the inside out using successive layers of burst, stars, and gampi paper. I think the third method might give the best symmetry, but it might be challenging to place the stars precisely inside the shell.

 

From a practical perspective, in my primitive attempts to make the effect, I've used thin rings of tissue paper, secured to the inside of the hemisphere. The stars can be placed inside these rings, and secured in place with burst charge. I've made the effect in three inch shells, and I've seen some Japanese shells of shells where the inserts must have been 2" or less. It's really important that the stars are primed well, since there are so few in the shell.

 

 

 

Regarding the ghost shells: A relay is probably used between color layers, but only for the purpose of making the color change consistent between the stars. The actual burn duration of the stars is generally dependent on the thickness of the layer of color composition. It's the thickness of the color layers that is varied to create the timing differences.

[wizard7611]

How can I synchronize burn times on stars?

[Mumbles]

Practice rolling stars until you're good at it. Star sizing screens are also helpful.

[wizard7611]

Why would you want this in a firework shell? I'm starting to see this in some Spanish shells and lots of Japanese shells.

 

http://i.imgur.com/kItZUQm.png

[zan89]

it is just a spacer to make room for next petal.

[wizard7611]

it is just a spacer to make room for next petal.

Ah I see. I assumed it was part of the effect since the video jumped to the shell hemispheres being pounded together.

[FloridaCracker]

I've heard they use clusters of tubes of paper to align the stars in a radial fashion. The size of the star and placement from the center also has to do with the placement while flying thru the air.

 

There are also dark primes where very little light is emitted that are used when stars all the sudden turn on or off or a wave of darkness flows across the break, etc. The Japanese are masters at rolling stars and supposedly roll the diameters and layer thicknesses to fractions of a mm to achieve many of the effects. I don't think I've ever seen another country come close to the precision they get. Plus they have amazing colors and perfect ignition. TR comes about the closest I've seen in the US but are quite different effects.

[wizard7611]

Okay, so what you're trying to say is they use paper tubes in-between comet stars like a checkerboard pattern in both hemispheres and fill the inside of the tubes with breaking powder?

 

Use this picture as a reference to what I'm trying to say :

 

http://www.skylighter.com/images/how_to/Stars/Fireworks-Hemisphere.jpg

 

So here you have a large shell(I would guess a 12 inch shell) with comets that are primed and hot-glued inside of it(never hot-glue stars!). Imagine paper tubes that are approximately one inch in diameter(inner diameter that is), .1 inches thick, and have the same height of the comets. Now, for every comet star placed inside of the shell, four of these tubes are placed surrounding the comet star, but glued in place. When gluing these tubes in place, all comet stars are extracted from the shell once the gluing is done. Once the gluing is finished, the comets are put into their positions and breaking powder is placed inside all of the tubes by tipping the shell hemispheres a bit to make sure it's evened out. I assume after those steps, a smaller version of this is accomplished, but this time smaller tubes would be used containing color stars and breaking powder would be used between the tubes as a replacement for the comet positions. After that's done, I assume the origin breaking charge is added as well as a booster. To be honest, I think my idea might actually work. I'll make a 12 inch shell someday and try it myself!

Edited December 4, 2015 by wizard7611

[burningRNX]

 

How can I synchronize burn times on stars?

Use compositions with the same burnrate, I believe beachle experimented with it and some other.

Ghost stars are hard to master without very precise star sizing (think about around 0,1 mm layer thickness for exact burnrate timing), and large batches I think, hard for a amateur, but a piece of cake for a large factory sadly

That snowflake again, I can not wrap my head around it exactly, what come's to mind is comets wrapped with kraft and back-filled with collor changing stars directed to the break, its the best way to give them that momentum and tight following symetry I think, but hey I'm no professional, Maybey someone is gonna give it a try

[Mumbles]

I've watched this particular shell several times now. If you watch it in slow motion, or pause several times you begin to notice a few things. I'm going to refer to the outer part as arms just for ease.

 

1. There are rising effects with multiple color changing stars. A given I think. These seem to have a few extra color changes than the stars in the main shell. I think I see blue to green to yellow to red to flash core.
2. There are clearly at least 2 separate petals. Again, I think this is a given, but included it anyway.
3. The inner petal is dark to white to strobe/shimmer.
4. If you watch it right after it bursts, it looks like the arms are one initial cluster.
5. As the shell burst grows, they begin to separate, particularly the silver streamer.
6. The color changing stars trailing the streamer look vaguely circular in some places.
7. I initially thought there were several units per cluster/arm, but now I think it's just how things were oriented with regard to the camera. Some of them just overlap in the field of view. There is a cluster at about 7 o'clock that looks to be alone, which makes me think they're all separately placed.

Overall, I don't believe this type of shell is overly complicated. Well, the complicated part is clearly the precision star rolling, not the actual assembly at least. This looks to be basically a kaliedoscope shell. Some stars are contained in tissue paper, and adhered to the shell wall casing in predetermined locations. Once all the bundles are attached, the shell is assembled like any other double petal. I'd bet the cluster that forms the arm looks something like the following:

 

[wizard7611]

Maybey someone is gonna give it a try

I'm going to try it.

[burningRNX]

After watching it several times, I considder:

 

After some counting the clusters of blue-green-red seems to be clusters of 7 stars. (an asymetrical number adds additional beauty imo)

 

The small groups of collor stars must lay between the OP and IP, and are in the same path as the silver streamer and b/g/r changers, I can not distinguish a lonely group of them and the count is almost the same for any larger cluster of systems, between sparse area's, (system = group of collor, 7 b/g/r changers and a streamer)

 

The lonely group at 7 o clock are two systems overlapping exactly, one streamer falling behind a little. (density in trail varies a litle showing a gap in expected tail density following the lowest streamer, pauze at the change between blue and green. )

 

 

If there are 7 b/g/r stars and 1 in the middle they are almost comparable in size making the silver streamer traveling almost the same path, why not using a larger comet? I hear you say, but when that happened the builder could arrange more changers around it withoud giving up the narrow dispersion cone. so why diddn't he?

 

I can not tell if the streamer was a comet or a round, but from the looks (I overlooked that is was a changer also) it should be a round.

A possibility is that the changers lay very close behind the streamer in the ring shape mumbles mentioned before with or without tissue wrap, and behind that the collor stars, I dont think that there is many room for seperated zone's with break charge between them because of the IP size.

[Mumbles]

I should have mentioned that my diagram didn't have the correct number of color changing stars. I was mostly going for general orientation and to demonstrate that it's probably not as complex as initially thought. I agree that there look to be 7. It was just easier to keep copy and pasting the first circle I made rather than try to appropriately size them for the correct number. I also fully agree that odd numbers just look better. It seems trivial, but it's the little things.

 

I know 7 stars will pack into a perfect hexagon with one in the center, so I agree that 7 around the periphery would be almost the same size. Do you think the streamer is just slower burning? Thus it's mass is conserved longer, and can separate more? My intuition says that this shouldn't be right though. Streamers tend to burn faster than colors if anything. This leads me to another thought. What if the streamer was significantly larger than the color changers? If this were the case, perhaps the color changing stars nest around the streamer, thought still around 7 stars. I've never used drastically different sized stars in the same petal, so I'm not sure how it affects the symmetry, and distance the stars are propelled from the burst. Using a pasted comet would increase mass, slow the burn rate, and be easy to pack around. Given how clean and crisp the color change from silver to red it however, I think it's rolled. It's at least not a cavity comet, though could be a very precise married comet.

 

With this regard, I feel it would be easier for the color changing stars to actually be exterior to the streamer. Trying to make a "crown" on each star and keep it oriented in the shell while loading and filling seems troublesome. It could be more of a pedestal of color changing stars with the streamer perched on top or at least sitting in the gap. The streamer should quickly make up any difference in initial velocity with it's higher momentum.

[burningRNX]

 

I know 7 stars will pack into a perfect hexagon with one in the center, so I agree that 7 around the periphery would be almost the same size. Do you think the streamer is just slower burning? Thus it's mass is conserved longer, and can separate more? My intuition says that this shouldn't be right though. Streamers tend to burn faster than colors if anything. This leads me to another thought. What if the streamer was significantly larger than the color changers? If this were the case, perhaps the color changing stars nest around the streamer, thought still around 7 stars. I've never used drastically different sized stars in the same petal, so I'm not sure how it affects the symmetry, and distance the stars are propelled from the burst. Using a pasted comet would increase mass, slow the burn rate, and be easy to pack around. Given how clean and crisp the color change from silver to red it however, I think it's rolled. It's at least not a cavity comet, though could be a very precise married comet.

I don't know if a larger star in the same petal will fly further out either, I never tested it, but its a key factor I agree

But otherwise stars that lay further away from the center seems to fly out the greatest distance, so that makes me believe that the collor changers lay further inside, probably directly behind the streamer witch could be any size then.

But then another problem arises, the radius of the dispersion cone will increase exponentionally when that stars lay further inside

 

Apart from all, the collor changing stars have to be extremely small, (when rolled) or there has to be another way to contain them in such narrow bundle, maybey still something with paper confinement,

Another explanation could be the extra breakcharge in the voids between the pattern that narrows it down.

 

Does anyone know how small a layer can be to give such duration?, personally I would go for around 2 mm thickness for a KP based composition.

[chonk]

How is this shell made(0:37)?

 

https://www.youtube.com/watch?v=SRVYagcmQnQ

 

Oh, I almost forgot. How do you roll ghost shell stars?

 

-Thanks!

Hi, I know I'm a bit late, but regarding the ghost shell, theres a rather famous fireworks factory in japan known for their ghost shells called "Marugo" or マルゴー. You probably have definitely of this company's shells somewhere on youtube. They make one of the the best ghost shells in my opinion in japan and I will be using one of their shells as a reference.

 

Their most basic ghost effect is the following:

https://youtu.be/Q4MAM9WZT0c?list=PLFzQvDb42TsSZs6c7-wlULG1WbEfcUbNd&t=123

(You don't have to click throughout the video to find the shell since I've already located the time stamp)

 

This is a 5" blue to yellow ghost shell.

Slowing it down you will notice there are six different segments or intervals where they change color. The shell is designed to make it look like the star is gone in a quick flash of yellow after burning blue.

This is done by using a very hot and bright and fast burning comp. (note that they can make them red, green, yellow, orange, and white)

 

However looking carefully, comparing the first segment to turn yellow to the last segment to turn yellow, the yellow core of the first segment burns significantly longer than the last segment, meaning its thicker.

They start off by making six different sizes of the yellow core, all of them being a constant interval torwards one another (.ie 1mm, so the cores would be 3,4,5,6,7,8mm), then coat it in a layer of blue until ALL of them are the same size.

 

The key is to have a slow burning comp turn into a fast burning comp. It is this that makes this company's shells so dramatic imo.

 

there is a long exposure on one of these exact shells but i will have to find it.

edit. found it. this these stars are in a ring pattern but are most likely made in the same manner as the one i described earlier. this shell is made by marugo as well.

 

 

or

 

Edited July 14, 2023 by chonk

[Edwin]

On 11/29/2015 at 12:04 PM, wizard7611 said:

Also, for the ghost effect, are color changing delay compositions used to create this effect?

It looks as if there is no relay at all due to no dimness in between the color changes. I'm assuming it's a very thin layer of step prime since those stars are probably of a very similar makeup, minus the flashcore (Perchlorate and mgal). 

This show is performed by marugo (a company that is rather close to the the company that made that kaleidescope shell (Saiki Enka)) in Thailand. 

Notice those color changing mines, those are 2" pillar mines and those stars look to be slightly larger than .5 in yet they change colors 7-8 times with no dimness in between the colors. Once again, looking like there is no changing relay whatsoever. 

What could've they possibly used?

Edited March 9 by Edwin

[Edwin]

On 12/7/2015 at 6:55 PM, Mumbles said:

I've watched this particular shell several times now. If you watch it in slow motion, or pause several times you begin to notice a few things. I'm going to refer to the outer part as arms just for ease.

 

1. There are rising effects with multiple color changing stars. A given I think. These seem to have a few extra color changes than the stars in the main shell. I think I see blue to green to yellow to red to flash core.
2. There are clearly at least 2 separate petals. Again, I think this is a given, but included it anyway.
3. The inner petal is dark to white to strobe/shimmer.
4. If you watch it right after it bursts, it looks like the arms are one initial cluster.
5. As the shell burst grows, they begin to separate, particularly the silver streamer.
6. The color changing stars trailing the streamer look vaguely circular in some places.
7. I initially thought there were several units per cluster/arm, but now I think it's just how things were oriented with regard to the camera. Some of them just overlap in the field of view. There is a cluster at about 7 o'clock that looks to be alone, which makes me think they're all separately placed.

Overall, I don't believe this type of shell is overly complicated. Well, the complicated part is clearly the precision star rolling, not the actual assembly at least. This looks to be basically a kaliedoscope shell. Some stars are contained in tissue paper, and adhered to the shell wall casing in predetermined locations. Once all the bundles are attached, the shell is assembled like any other double petal. I'd bet the cluster that forms the arm looks something like the following:

 

Looking from the amount of clusters there are, if these clusters are made laterally like a biscuit, there wouldn't be much space to fit all those clusters in that shell.

My guess is porforated hemis like so:

Source: Facebook (Akihiro Koga) 

[Zumber]

3 hours ago, Edwin said:

 

 

Notice those color changing mines, those are 2" pillar mines and those stars look to be slightly larger than .5 in yet they change colors 7-8 times with no dimness in between the colors. Once again, looking like there is no changing relay whatsoever. 

What could've they possibly used?

For pillar ( pumped star/ comet) Changing relay is not necessary rather sequence of stars composition should be capable of igniting next layer of colour composition.

In encyclopedic dictionary I have seen kraft paper tube is used where three or more colour composition of desired thickness is pressed and at the end prime layer is applied, this technique is useful in mines but for round stars changing relay is necessary.

[Crazy Swede]

I’m not sure if there still are open questions in this thread but kaleidoscope shells are made with hemispheres as in the picture above. There has been evidence (found on the ground in Japan) that they also can stack smaller hemispheres on top of the bigger ones. 

Regarding dark relay formula, a friend of mine developed a very effective composition based on potassium nitrate and the type of phenol/formaldehyde resin powder that contains hexamine for curing. I do not know the exact ratio but it was a very fuel lean composition and it was rolled on to perfectly round stars, that was sorted in size with 0,1 mm accuracy, in a very thin layer. Probably less than 1 mm.

[Zumber]

1 hour ago, Crazy Swede said:

I’m not sure if there still are open questions in this thread but kaleidoscope shells are made with hemispheres as in the picture above. There has been evidence (found on the ground in Japan) that they also can stack smaller hemispheres on top of the bigger ones. 

Regarding dark relay formula, a friend of mine developed a very effective composition based on potassium nitrate and the type of phenol/formaldehyde resin powder that contains hexamine for curing. I do not know the exact ratio but it was a very fuel lean composition and it was rolled on to perfectly round stars, that was sorted in size with 0,1 mm accuracy, in a very thin layer. Probably less than 1 mm.

It would be better if he is willing to share dark relay formula.

And video of shells with this developed changing rely if he has.

Dr takeo shimizu also studied kno3 based changing relay.

Kno3 84

Charcoal paulowinia 08

Sulphur 08

Binder 2 to 3 percent.

Present formula have positive oxygen value which is necessary for changing relay & amount of binder is very important, formula with negative oxygen value is used for time fuse.

[Edwin]

3 hours ago, Crazy Swede said:

I’m not sure if there still are open questions in this thread but kaleidoscope shells are made with hemispheres as in the picture above. There has been evidence (found on the ground in Japan) that they also can stack smaller hemispheres on top of the bigger ones. 

Regarding dark relay formula, a friend of mine developed a very effective composition based on potassium nitrate and the type of phenol/formaldehyde resin powder that contains hexamine for curing. I do not know the exact ratio but it was a very fuel lean composition and it was rolled on to perfectly round stars, that was sorted in size with 0,1 mm accuracy, in a very thin layer. Probably less than 1 mm.

That picture was from a Japanese fireworker. Btw I never heard or seen of the hemisphere stacking technique. 

Edited March 9 by Edwin