http://www.math.utoledo.edu/~codentha/Cryptanalysis/lanaki/lanaki24.txt

UBCHI

The Ubchi (the U is umlauted) is a double columnar
transposition cipher used by the Germans during WWI. It
was broken by the French thanks to in part to a radio
message sent in unprotected cleartext early in the
conflict.

The Ubchi had a keyphrase that was represented by
numerals according to the position of its letters. Two
identical letters were labeled consecutively if they
appeared in the same keyphrase. For example,

5 3 7 8 9 2 6 1 4 10
Keyword:    h e r r s c h a f t

For the plaintext: First army X Plan five activated X
Cross Marne at set hour.

Ciphertext key block 1:

5 3 7 8 9 2 6 1 4 10
h e r r s c h a f t
——————-
F I R S T A R M Y X
P L A N F I V E A C
T I V A T E D X C R
O S S M A R N E A T
S E T H O U R

The ciphertext was taken off by columns in numerical
order of the keyword columns:

1     2     3    4     5     6     7
Ciphertext:  MEXE AIERU ILISE YACA FPTOS RVDNR RAVST

8     9    10
SNAMH TFTAO XCRT.

(Note the 5 letters groups not observed.)

These groups were then transcribed horizontally into
another block beneath the same number sequence:

5 3 7 8 9 2 6 1 4 10
h e r r s c h a f t
——————-
M E X E A I E R U I
L I S E Y A C A F P
T O S R V D N R R A
V S T S N A M H T F
T A O X C R T(Z)

The next step was to add as many Null letters as there
are words in the Keyphrase or Keyword. One null Z was
added after the last letter in the last row, T.

The German encipherer once more took these letters from
the block by columns in the same numerical sequence and
separated into standard groups of five letters each:

1     2    3      4     5    6      7    8     9
RARHZ IADAR EIOSA UFRTM LTVTE CNMTX SSTOE ERSXA YVNCI

10
PAF.

To decipher the message, the recipient first had to
discern the size of the transposition rectangle in order
to learn how long the columns were. This was accomp-
lished by dividing the total number of key numbers into
the total number of letters into the message (48 / 10).
The quotient was the number of complete rows. The
remainder 8 was the number of letters in the incomplete
columns.  The succeeding steps reversed the corre-
sponding steps in the enciphering process.

Note the similarity with the U.S. Army Double Trans-
position Cipher System. Barker gives a detailed
breakdown of this type of cipher in his book.

It is not coincidental that the two countries at war
had very similar cipher systems in play.

Ubchi is nice because it’s a complicated cipher.  As a transposition, it is not enough on its own to solve K4.  Adding null letters to indicate keyphrase length is not ideal for what I have in mind for Kryptos personally but if you’re willing to accept a shorter K4 message then it is of little consideration.

You’ll need the keyword to allow the manipulations.

I tried Kryptos:
OFQI
OGJKQIKIWJBXWLBBQNVRYSOTOVKHIAS

Hydra:
PORM
QOKSLKZLJTBGURUWPKBKOVTUZSGWSBZ
ZICJWKDVAIRCNSBTWOFJHGESXWFDKFU
IQUNNAYSKTTXPQUDKOTABILLHAQEFG

and Tutankhamen:
XMAZ
WCDBWUKIZKHLRTBEKOQROXVIRFDBASU