W.W.&F. Discussion Forum
April 25, 2017, 06:39:42 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
Notice: Great things in store at the WW&F.
 
   Home   Help Search Login Register  
Pages: 1 2 3 [4]
  Print  
Author Topic: Boiler plate flanging machine  (Read 5896 times)
Dave Crow
Museum Member
Hostler
***
Posts: 286


View Profile
« Reply #45 on: April 20, 2017, 08:16:49 AM »

Harold, it appears your patterns for the dies were a large part of the reason the formed sheet looks so good!  Congratulations.

Dave Crow
Logged
Brendan Barry
Museum Member
Engineer
****
Posts: 637


View Profile
« Reply #46 on: April 21, 2017, 11:00:47 AM »

The machine was put to work yesterday flanging one of no. 10's firebox sheets.























Logged
Rick Sisson
Museum Member
Gandy Dancer
*
Posts: 14


View Profile
« Reply #47 on: April 21, 2017, 05:15:48 PM »

Can we get a report from Harold, Eric, and Jason on how well the flanging process is going? Any surprises?
How much time is devoted to preparation (layout, machine setup, etc.) vs. actual flanging?
Also, can you describe the die(s) that are used to form those beautiful corner bends?
Thanks.
Logged
Jason M Lamontagne
Operating Volunteers
Conductor
*****
Posts: 821


View Profile
« Reply #48 on: April 21, 2017, 07:14:35 PM »

I'll try; Harold and Eric can chime in with their thoughts to if they wish.

Careful layout is important (somewhat obvious I suppose).  First, the exact shape of the desired flange is drawn on the sheet. In this case, it is traced from the one piece firebox, as the actual firebox shape is a little different than the print.  This final flange shape is drawn as the inside surface of the flange.  Once that's done, the centers of rotation are determined for the curved areas.  We then pick a standard offset from the flange line.  For straight areas, the offset line is parallel.  For curved areas, the standard offset is swung as an arc off the center of rotation.  Points of tangency, centerljnes, and sometimes another bisecting set of lines, are drawn on both the flange edge and the offset line.

The pointer on the flanger anvil is set at the offset distance from the face of the die being used. 

The toggle is brought down to 90 degrees, the anvil brought up to one sheet thickness away from the toggle shoe face, and the backside of the anvil blocked to that location off the anvil stop.  That's 0 location for that operation.  The anvil is then blocked either in by 3", and the plate bent about 20 degrees at each index mark for that operation.  The anvil is then backed off to 0 + 2" and bent to about 45 degrees.  On the easier curves, we then back off to 0 + 1" and bend to about 75 degrees.  On harder bends, adjacent areas must be bent first. 

As expected, some buckling develops on the sharp corners.  Here's the cool part:  the flanger's available 200 tons exerted on the flange simply smushed those bumps flat.  Pretty much like they never existed. 

The flanger never maxed out, but does exert more effort when bending more plate on a curve, against adjacent areas which aren't flanged as much.  Again, expected. 

We're achieving exact knuckle radii and flange locations within 1/32". 

Our indexing method of lining up index prick punches to the pointer, under the plate, is a little tedious, but is a very direct method.  We're considering going to a template and follower system for flanged plates requiring tighter tolerances. 

It's been a lot of fun finally seeing this machine in use.  I've had it in mind since 2006, when we hot formed No 9's flanged boiler plates. 

See ya
Jason
Logged
Jason M Lamontagne
Operating Volunteers
Conductor
*****
Posts: 821


View Profile
« Reply #49 on: April 21, 2017, 07:24:40 PM »

After roughly describing the layout method, Harold actually performed all the layout work.  Very well executed, and refined as he went along. 

We've found it best to use a die which matches the shell radius for each area.  It's possible to use one with more radius, but aligning the plate for subsequent hits is much harder, due to undesired bend start locations on either side of center from the hit using an incorrect die. 

The nice corners are a 6" shell radius, formed on a die with the same radius.

Harold predicted this detail and made patterns for all of 10's various shell diameters on the flanged plates.

We haven't really experimented yet with using dies that have a close shell radius.  That will determine if we can, for example, use 10's 6" radius firebox corner die on 11's 8" radius firebox corners.

See ya
Jason
Logged
John Kokas
Museum Member
Engineer
****
Posts: 598


View Profile
« Reply #50 on: April 21, 2017, 07:50:27 PM »

The results look fabulous.  Keep up the good work guys!  A side note question; as we get more adept at this process is there any possibility for future outside contract work?
Logged
Harold Downey
Museum Member
Switchman
**
Posts: 64


View Profile
« Reply #51 on: April 21, 2017, 08:34:54 PM »

Some more details:

I probably spent an hour doing the layout on the rear door sheet (back inside of firebox), since it has 2 corners with 6" radius, top has 31" radius, sides have two straight sections separated by another small bit of 31" radius.   Then after some learning,  I needed to add more layout lines in the 6" corners, and also color code all the index marks, since it got very confusing which mark to use for a given bend operation.      Simpler parts take a lot less layout time. 

I overshot the mark a bit in making anvil dies and toggle dies.   I made a matching toggle die for each anvil die.  Up to now we have only needed a flat die on the toggle and the correct radius die on the anvil.  So that was good news for future projects.    I don't yet know what happens for inside radius situations like the throat sheet.  A radius die on the toggle is definitely needed, but is a flat die on the anvil good enough?   

We're still learning, and each sheet is unique in its own way.   We are taking about four steps for each bend, and sometimes another round to smooth out small bulges or imperfections. 

I'm looking forward to doing the backhead.  It has much longer flanges (for riveting), and really big bend radii, compared to the door sheet.   We may need to suspend it from the overhead crane, because it is substantially heavier too. 

The flanger is working marvelously.  It is yet another example of the WW&F spirit.   It was an ambitious project with more than a bit of risk, and it took a combination of many volunteer skills to carry out.  Kudos to Jason for the concept! 

Harold
Logged
John Scott
Museum Member
Flagman
*
Posts: 39


View Profile
« Reply #52 on: April 21, 2017, 09:17:26 PM »

Impressive skills are on display, here.
Logged
John McNamara
Operating Volunteers
Dispatcher
*****
Posts: 1,005


View Profile
« Reply #53 on: April 21, 2017, 09:40:43 PM »


As expected, some buckling develops on the sharp corners.  Here's the cool part:  the flanger's available 200 tons exerted on the flange simply smushed those bumps flat.  Pretty much like they never existed.  

{intervening text deleted}

We're achieving exact knuckle radii and flange locations within 1/32".

Is "smushing" legal in Maine?
I can remember many years ago, you and I used to joke that 1/32" was standard track crew tolerance. Grin

-John M
Logged
Jason M Lamontagne
Operating Volunteers
Conductor
*****
Posts: 821


View Profile
« Reply #54 on: April 21, 2017, 10:27:12 PM »

Maybe it should have said "smooshing?"

Harold says tolerance for settings is +/- 0.000.  He's got me beat!

See ya
Jason
Logged
Roger Cole
Flagman
*
Posts: 48


View Profile
« Reply #55 on: April 22, 2017, 05:15:05 PM »

What is the thickness of the steel plate you were flanging?
Logged
Harold Downey
Museum Member
Switchman
**
Posts: 64


View Profile
« Reply #56 on: April 22, 2017, 05:39:18 PM »

3/8" - true for all of #10's sheets.   One sheet for #11 is 1/2"
Logged
Pages: 1 2 3 [4]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.21 | SMF © 2006-2008, Simple Machines Valid XHTML 1.0! Valid CSS!