Following Julius Hoffman's advice. The Hoffman edge

In 1879 the design of the Hoffman Frame was announced to a wide audience through a letter written to The Bee Keepers Exchange (article link). For many months/years after this letter, beekeepers from across the US wanted to add their own thoughts and modifications to the frame and provided this feedback by writing in to publications such as ‘Gleanings in Bee Culture’. These letters appear to have been a point of frustration for Julius Hoffman, leading to the letter below from Julius Hoffman himself, ‘clearing up’ some of the misconceptions and criticisms of his frame design.

The article clarifies several items:

  1. The frames end bars should be beveled on one side
  2. The constant width of the top bar between the end bar and the side of the box is intentional to cover the rebate/ledge in the box
  3. The radius pictured on the end of the top bar is to provide a pivot/hinge for the frame and is intentional
  4. The direction of the Hoffman edge is to have beveled edges pointing clockwise around the box which is shown in the diagram of adjacent frames (assuming the image is a view of the top of the frames). This is important for consistency across hives (ie. to avoid two pointed or flat edges contacting)
  5. The wide section of the end bar should extend at least 3 inches down from the top bar (this is still honoured in current frame designs)

This forum post really focuses on item 1., the beveled edge, and reimplementing this feature on end bars, 127 years after Hoffman identified the issue with beekeepers not correctly following his design.

The full letter transcribed below is from Gleanings In Bee Culture, Volume 18, Page 489 (1890)

What the inventor says of it.
In reading No. 8,9,10, and 11 of Gleanings I find that, in discussing “spaced frames”, the frames I invented and used for about sixteen years have been brought to notice by some of our bee-keeping friends. But as the description and illustrations of the frame as I have them lack some important points I will make a few remarks about them. One of the edges of the uprights, or ends of the frame, as far as they are close fitting, should be beveled off to a little less than 1/8 inch, so as to meet at the center of the square edge of the next frame, as in E, of the diagram. This will prevent the gluing and sticking together of the frames, also squeezing of the bees to a large extent, and allow faster working them.

The concerns of the top-bars of the frame at C should not be taken off more than just enough to allow for turning the frames easily in the rabbets, otherwise the bees will have too much of a chance to fill up the rabbets with propolis. For the same reason, I object to your modifications of the top bars as the jutting points of them, as called by Ernest, are very important, in order to fill out the rabbets to keep them free from bee-glue. Yo use these frames on tin rabbets, as proposed by Ernest, would, according to my experience, be a very doubtful improvement. The frames would slide too easily on the rabbets and the carrying and moving the hives would soon convince him of it. The way I use them on wooden rabbets, the frames are just glued fast enough to allow the moving of the hives without any preparatory fastening, unless the hives are only partly filled with frames. In this case I take a piece of burlap, or any cloth that I use for coverying up the frames; roll it into a bunch and stuff it behind the spacing-board.

The close fitting part of the side-bars should extend at least 3 inches down from the top-bar. As a well-made frame lasts a good many years, their extra expense amounts to but very little, if they have any advantages over others.

JULIUS HOFFMAN. Canajoharie, N. Y., June 15, 1890.

Taking Hoffman’s measurement and adapting it to a Hoffman end bar for a Langstroth frame, the angle of the cut could be calculated. From the centre ‘peak’ in the frame, it’s roughly a 32 degree angle. The following diagram shows the ‘waste’ material on the end bar as blue:

Note: The angle in the drawing provided by Hoffman and the CAD drawing below look significantly different, I believe this is due to Hoffman exaggerating the angle in his drawings to emphasise the feature of the end bar.

Rather than using a standard 45 degree chamfer bit, a 30 degree chamfer bit was found to more closely match the original angle. The router bit is manufactured by a tool company in Victoria, Australia (Carb I Tool)

The next stage required the convincing of a local woodworker to exchange time on a router table for honey. Luckily the last season in Canberra provided enough honey to route a full box of frame ends.

The 30 degree chamfer router bit was fitted in a fixed table router and the fence was adjusted over several iterations to correctly locate the centre of the frame edge.

An off cut from the bottom of an end bar was used to hold the end bar square through the cut.
The end bars were then cut in two passes down the same edge to form the bevel described by Hoffman.
It was soon identified that the inconsistency in the frame thickness (+/- 0.5mm) resulted in the routed edge not always forming a perfect point. This is shown in the photos below, where there is a very minor flat edge left remaining on some of the thicker end bars. Leaving this flat edge was chosen over the alternative adjustment that may have modified the 35mm width of the frame if the cuts were closer to centre.
Before and after comparison showing the beveled edge of the end bar.
The next step is to assemble these frames as described by Hoffman, and run them for the 2017/2018 season to confirm the angle works correctly for its intended purpose of limiting propolis build up in the contact points between adjacent end bars.

These modified end bars are now assembled in frames with the Hoffman edges pointing in a clockwise direction.

One thing that became immediately apparent is the reduction in friction when pulling tightly packed adjacent frames. Even with the minor inconsistencies in the bevel due to end bar thickness, the overall improvement is considerable.

The following shows the frames pressed tightly together and the Hoffman edge maintaining correct spacing with limited contact area.

When the frame are pressed tightly together, you can see the 'gap' the is introduced by the bevel on the face of the end bars.

It's unfortunate that several small features such as this have been discarded in the push for automation/mass production of bee frames, especially when the impact to the handling/manipulation of the frames is so noticeable.

With the success of these changes the next step is to do the same on some Ideal depth frames in preparation for a full season of testing with a hive containing only Hoffman end bars.

One small issue I have identified is that the Hoffman edge (when pointed clockwise) is problematic in my current wiring jig as this is the side that nails are placed in. I will post some photos of wiring these frames in my updated jig over the next week or so.

Happy to report that the first batch of Hoffman edge ideal end bars are completed and work just as well as the full depth equivalents.

Next issue to overcome is wiring of these frames as the beveled edge is on the side (right hand side in below image) that I usually fasten the wire to and this is also the orientation my frame wiring jigs are configured for.

Minor update. I have the Ideal frames wired and ‘checker boarded’ them through last years supers which I started adding back to hives over the last week or so. Even between the standard end bars the edge appears to be make a difference.

It will be interesting if any difference is still noticeable after a solid season of propolisation!

I found this image in the 1907 Dadant revision of Langstroth’s ‘The Hive and the Honey Bee’ on page 150 referencing an ‘ABC of Bee Culture’ image showing the Hoffman edge in an anti-clockwise (counter-clockwise if you’re American) configuration. I’ll still take the original creators drawing… but interesting to see the alternative published: