Login    Register
User Information
Username:
Password:
We are a free and open
community, all are welcome.
Click here to Register
Sponsored
Who is online

In total there are 58 users online :: 4 registered, 0 hidden and 54 guests


Most users ever online was 155 on Mon Aug 15, 2016 1:40 am

Registered users: Bing [Bot], Google [Bot], MSNbot Media, Yahoo [Bot] based on users active over the past 5 minutes

The Team
Administrators
Global Moderators
global_moderators.png CS

HGDT 0.4 online

Post questions and info about hybrid (compressed gas with fuel) powered cannons here. This includes discussion about fuels, ratios, ignition systems, build types, safety, and anything else relevant.
Sponsored 
  • Author
    Message

Unread postAuthor: jimmy101 » Sat May 03, 2008 3:21 pm

I tried modelling my standard 1x gun. 11x3" chamber, 30"x2" barrel, 100g spud, double bevelled spud knife, WAG for static friction 30 pounds (used to set the burst disk pressure at 10 PSI). WAG dynamic at 1/2 that (used 5 PSI as the "friction" value). "Turbulance induction" checked, single spark, 1X, HGDT version 0.4.2 (one behind most recent)
...
gives muzzle velocity of 229 FPS.

Gun actually fires at 300~330 FPS. Tweaking the parameters that I'm not certain of (primarily the static friction and the mass), I can't get the velocity up over 300 FPS.


With a "mine sized" chamber the problem isn't heat loss, it is how long it takes the chamber to burn and what the ammo is doing during that time. A very large chamber burns very slowly. So slowly that the projectile transits the barrel at an essentially constant pressure determined by the static friction between the ammo and barrel. A high pressure burst disk circumvents this problem.

Nobody knows how big "mine sized" is. My guess is that it depends on many factors (like pretty much everything else in combustion spud guns). For a moderately high friction ammo, I would expect the onset of "mine sized" to be anything with a CB greater than 1.5 to 3 and more. In other words, a plot of velocity versus chamber volume for a fixed barrel will peak somewhere in the CB range of 0.8 to 3 and drop after that.
  • 0

Image

jimmy101
Lieutenant General
Lieutenant General
 
Posts: 3128
Joined: Wed Mar 28, 2007 9:48 am
Location: Greenwood, Indiana
Country: United States (us)
Reputation: 7

Unread postAuthor: D_Hall » Sat May 03, 2008 3:38 pm

jimmy101 wrote:I tried modelling my standard 1x gun. 11x3" chamber, 30"x2" barrel, 100g spud, double bevelled spud knife, WAG for static friction 30 pounds (used to set the burst disk pressure at 10 PSI). WAG dynamic at 1/2 that (used 5 PSI as the "friction" value). "Turbulance induction" checked, single spark, 1X, HGDT version 0.4.2 (one behind most recent)
...
gives muzzle velocity of 229 FPS.

Gun actually fires at 300~330 FPS. Tweaking the parameters that I'm not certain of (primarily the static friction and the mass), I can't get the velocity up over 300 FPS.

Thank you for the data point. If I had to guess, I'd say my allowances for turbulence aren't big enough. That's one part of the puzzle that seems to have a large effect but is Damned Near Impossible to nail down in a generic manner as it's going to vary from gun to gun (how big is YOUR chamber fan compared to the next guy's?).



With a "mine sized" chamber the problem isn't heat loss, it is how long it takes the chamber to burn and what the ammo is doing during that time. A very large chamber burns very slowly. So slowly that the projectile transits the barrel at an essentially constant pressure determined by the static friction between the ammo and barrel. A high pressure burst disk circumvents this problem.

If I follow what you're trying to say... Yes and no. For "small mines" this is a valid statement. But for "large mines" you're going to experience flame front acceleration. You know, that thing that (if left unchecked and given enough room) leads to DDT. As the flame front accelerates... Well, what started out as a slow burn becomes Damned Fast.

Now, you can argue about whether or not I've implimented this well, but I *HAVE* implimented it.
  • 0

Simulation geek (GGDT / HGDT) and designer of Vera.
User avatar
D_Hall
Donating Member
Donating Member
 
Posts: 1759
Joined: Thu Feb 07, 2008 7:37 pm
Location: SoCal
Reputation: 6

Unread postAuthor: SpudFarm » Sat May 03, 2008 4:03 pm

hey on my hybrid it is a "bit" optimistic..

it says 430m/s (about) and i have chronyed it to 327m/s

that is a pretty big change there..
  • 0

"Made in France"
- A spud gun insurance.
User avatar
SpudFarm
Donating Member
Donating Member
 
Posts: 2563
Joined: Sat Nov 04, 2006 9:39 am
Location: Norway Trondheim area
Reputation: 0

Unread postAuthor: D_Hall » Sat May 03, 2008 4:20 pm

spudfarm wrote:hey on my hybrid it is a "bit" optimistic..

it says 430m/s (about) and i have chronyed it to 327m/s

that is a pretty big change there..

Well, what can you tell me about your hybrid?
  • 0

Simulation geek (GGDT / HGDT) and designer of Vera.
User avatar
D_Hall
Donating Member
Donating Member
 
Posts: 1759
Joined: Thu Feb 07, 2008 7:37 pm
Location: SoCal
Reputation: 6

Unread postAuthor: SpudFarm » Sat May 03, 2008 4:25 pm

2" by 12" chamber
0.5" by 80" barrel
5x
one spark
burst disk that don't burst even after two shots of the same disk at 4x

more needed?

Edit: the projectile was 5gram and the bore 1.6cm (i am metric)
  • 0

"Made in France"
- A spud gun insurance.
User avatar
SpudFarm
Donating Member
Donating Member
 
Posts: 2563
Joined: Sat Nov 04, 2006 9:39 am
Location: Norway Trondheim area
Reputation: 0

Unread postAuthor: jimmy101 » Sat May 03, 2008 5:13 pm

D_Hall wrote:For "small mines" this is a valid statement. But for "large mines" you're going to experience flame front acceleration. You know, that thing that (if left unchecked and given enough room) leads to DDT. As the flame front accelerates... Well, what started out as a slow burn becomes Damned Fast.

Now, you can argue about whether or not I've implimented this well, but I *HAVE* implimented it.

I think I understand what you are saying, flame front acceleration beyond what the simple power law predicts, the Y in;

S<sub>flame</sub>=S<sub>flame0</sub>(P<sub>i</sub>/P<sub>0</sub>)<sup>a</sup>(T<sub>i</sub>/T<sub>0</sub>)<sup>b</sup> + Y

But, wouldn't any flame front acceleration above the power law occur after a fair amount of the chamber has burned? For static frictions in the range of typical spudguns this may well occur after the spud has already left the barrel. 30 pounds static force (friction) is just 10 PSIG in a 2" barrel. Propane + air combustion will reach that pressure when just (roughly) 13% of the fuel has burned. For "typical" spud + barrel + velocities the transit time is only 10~30mSec. I wouldn't think that there is much difference between what a large chamber was doing at 13% combustion and what it will be doing 10~30mSec later. Obviosuly though, this is the kind of a question where "God is in the details". For some conditions it might not matter, other times it might.

I ran the same gun through Evbec Live v1.5. Evbec gives a muzzle velocity of 226 FPS, the same as HGDT's 229 FPS. Evbec doesn't have friction as an input parameter so there is no way to model the difference between a low friction round and a high friction one, such as the difference between a single bevel and a double bevel spudcutter. My gun has a double bevel cutter and it takes a lot of force (as in leaning hard on the ramrod) to load a spud.

Evbec predicts that the velocity will turn over as the chamber volume is increased. According to Evebec the turnover CB is about 10:1. I don;'t believe anything was done in Evbec to predict this performance charachteristic. It just "falls out of the math". The CB ratio is higher than I would guess and may well be so far outside the range of practical spudguns as to be ignorable.
  • 0

Image

jimmy101
Lieutenant General
Lieutenant General
 
Posts: 3128
Joined: Wed Mar 28, 2007 9:48 am
Location: Greenwood, Indiana
Country: United States (us)
Reputation: 7

Unread postAuthor: D_Hall » Sat May 03, 2008 6:09 pm

jimmy101 wrote:I think I understand what you are saying, flame front acceleration beyond what the simple power law predicts, the Y in;

S<sub>flame</sub>=S<sub>flame0</sub>(P<sub>i</sub>/P<sub>0</sub>)<sup>a</sup>(T<sub>i</sub>/T<sub>0</sub>)<sup>b</sup> + Y

Yes and no.

The power law assumes a laminar flame front. In an ullage explosion (and that's basically what we're dealing with), a flame front will transition to turbulent very quickly and it will do so even in quiescent air and without having to interact with any vessel walls. When this happens, the ol' power law goes right out the window. So really what you end up with is a whole new equation rather than the Y component of the equation you list.

That said, I model it just like you imply... as the Y term. Given that I'm not really certain WHEN the transition applies and such, I figure a simple superposition is as good as any approach.

Note: I've seen high speed photography of ullage explosions as a result of a job I once worked. I know the transition to turbulent was certainly measured in ms. Unfortunately, I no longer have access to those pictures and can't seem to find any similar pics on google.

Obviosuly though, this is the kind of a question where "God is in the details".

Indeed.


By the way... The best/easiest way to model your gun in HGDT would be to put the disc failure at 0 psi and the friction to 10 psi. The friction input is, after all, the static friction with dynamic friction assumed to be 50% of static.
  • 0

Simulation geek (GGDT / HGDT) and designer of Vera.
User avatar
D_Hall
Donating Member
Donating Member
 
Posts: 1759
Joined: Thu Feb 07, 2008 7:37 pm
Location: SoCal
Reputation: 6

Sponsored

Sponsor
 


Unread postAuthor: D_Hall » Sat May 03, 2008 6:35 pm

Jimmy,

Quick question.... What's your chronograph set up? Do you have a shadow box set up or are you one of the guys running an acoustic system?
  • 0

Simulation geek (GGDT / HGDT) and designer of Vera.
User avatar
D_Hall
Donating Member
Donating Member
 
Posts: 1759
Joined: Thu Feb 07, 2008 7:37 pm
Location: SoCal
Reputation: 6

Unread postAuthor: D_Hall » Sat May 03, 2008 10:16 pm

spudfarm wrote:2" by 12" chamber
0.5" by 80" barrel
5x
one spark
burst disk that don't burst even after two shots of the same disk at 4x

more needed?

Edit: the projectile was 5gram and the bore 1.6cm (i am metric)


Friction?

But mostly though... Was your bore 1.6 cm or 0.5" It can't be both.... And are there any other dimensional oops in there?
  • 0

Simulation geek (GGDT / HGDT) and designer of Vera.
User avatar
D_Hall
Donating Member
Donating Member
 
Posts: 1759
Joined: Thu Feb 07, 2008 7:37 pm
Location: SoCal
Reputation: 6

Unread postAuthor: D_Hall » Sat May 03, 2008 10:46 pm

And in other news....

Don't know what to tell Jimmy or Spudfarm but....

I just made a slightly more intelligent guesstimate as to the effects of a chamber fan. Previously I'd just pulled "velocity delta" out of me arse. Today it dawned on me that the speed of the air driven by the fan might make an intelligent choice. I also corrected a bonehead on the heat transfer model (yes, Jimmy, there was an error that I've known about for about a week, but it isn't/wasn't enough to account for the things we're seeing). In any event, after doing that I was shocked....
  • 0

Attachments
latke_hgdt.gif
latke_hgdt.gif (10.76 KiB) Viewed 569 times
Simulation geek (GGDT / HGDT) and designer of Vera.
User avatar
D_Hall
Donating Member
Donating Member
 
Posts: 1759
Joined: Thu Feb 07, 2008 7:37 pm
Location: SoCal
Reputation: 6

Unread postAuthor: jimmy101 » Sun May 04, 2008 12:48 pm

D_Hall wrote:Jimmy,

Quick question.... What's your chronograph set up? Do you have a shadow box set up or are you one of the guys running an acoustic system?

<a href="http://www.inpharmix.com/jps/Jims_chrono.html">Homemade, phototransistors + PC soundcard</a>. Should be at least as accurate as a Chrony (assuming no stupid errors in the math).
  • 0

Image

jimmy101
Lieutenant General
Lieutenant General
 
Posts: 3128
Joined: Wed Mar 28, 2007 9:48 am
Location: Greenwood, Indiana
Country: United States (us)
Reputation: 7

Unread postAuthor: jimmy101 » Sun May 04, 2008 12:55 pm

Dave

So it looks like HGDT is hitting Latke's data perfectly (throw on Latke's error bars and there is little difference between the two sets of numbers).

IIRC, the two data points for spuds from the two shortest barrels are suspect anyway, so HGDT is probably doing better than it looks for spuds from a short barrel.

What did you estimate the mass of Latke's wood shell at? (IIRC it doesn't appear on any of his pages.)
  • 0

Image

jimmy101
Lieutenant General
Lieutenant General
 
Posts: 3128
Joined: Wed Mar 28, 2007 9:48 am
Location: Greenwood, Indiana
Country: United States (us)
Reputation: 7

Unread postAuthor: D_Hall » Sun May 04, 2008 1:05 pm

jimmy101 wrote:What did you estimate the mass of Latke's wood shell at? (IIRC it doesn't appear on any of his pages.)

1.2 oz.

I just guestimated dimensions from the photos and assumed a density appropriate for maple (or something that seemed appropriate at the time; don't actually recall).
  • 0

Simulation geek (GGDT / HGDT) and designer of Vera.
User avatar
D_Hall
Donating Member
Donating Member
 
Posts: 1759
Joined: Thu Feb 07, 2008 7:37 pm
Location: SoCal
Reputation: 6

Unread postAuthor: D_Hall » Sun May 04, 2008 5:27 pm

A question for the masses....

Not that it's horribly surprising, but it appears that chamber fan performance may very well have a significant effect on system performance. Currently, HGDT simply has a "yes/no" option for chamber fans (turbulence induction). I've decided that this is insufficient.

The question then becomes precisely what IS sufficient?

Should I have a pulldown box that allows you to select from none, low, med, high, extreme?

Or should I have a text box that allows you to manually input a burn rate delta-V?

The first would be nice because it's simple for the user and includes a few options. But it does open the door for "low is too low, med is too high" scenarios.

The second allows one to really dial in on a gun, but it requires a bit more smarts from the user.

A third option would be a whole 'nuther input section dedicated to the chamber fan....


Thoughts?
  • 0

Simulation geek (GGDT / HGDT) and designer of Vera.
User avatar
D_Hall
Donating Member
Donating Member
 
Posts: 1759
Joined: Thu Feb 07, 2008 7:37 pm
Location: SoCal
Reputation: 6

Unread postAuthor: SpudFarm » Sun May 04, 2008 5:37 pm

D_Hall wrote:
spudfarm wrote:2" by 12" chamber
0.5" by 80" barrel
5x
one spark
burst disk that don't burst even after two shots of the same disk at 4x

more needed?

Edit: the projectile was 5gram and the bore 1.6cm (i am metric)


Friction?

But mostly though... Was your bore 1.6 cm or 0.5" It can't be both.... And are there any other dimensional oops in there?


no..
the projectile was a marble "1.6cm" and i don't think that has much friction at all. th "oops" was made when i thought a marble was 0.5" ans then i realised that .5" is 1.27cm
  • 0

"Made in France"
- A spud gun insurance.
User avatar
SpudFarm
Donating Member
Donating Member
 
Posts: 2563
Joined: Sat Nov 04, 2006 9:39 am
Location: Norway Trondheim area
Reputation: 0

PreviousNext

Return to Hybrid Cannon Discussion

Who is online

Registered users: Bing [Bot], Google [Bot], MSNbot Media, Yahoo [Bot]

Reputation System ©'