Yes, there are some standard setups/patterns that are useful specifically for defense, and the most "absolute" such defensive setups in chess go under the term fortress, which are setups in which the stronger side is provably unable to make any progress toward a win. Valentin's answer, for instance, already gives a nice example of a known fortress, wherein the white king stays in the corner and the bishop and knight coordinate to prevent Black from closing in and delivering mate.
As Valentin points out, though, his is a rather uncommon example in practice. The article I linked to above gives a nice example of a sort of piece-down fortress that arises somewhat more often, and can be worth remembering:
[FEN "8/1b6/8/8/8/6p1/3k2P1/5K2 w - - 0 1"]
1.Kg1 Ke2 2.Kh1 Kf2
The white king stays in the corner, and Black can't threaten the g-pawn without stalemating. Note too that the color of the bishop doesn't matter either; a dark-square bishop would of course be drawn too. Again, the linked article gives a fantastic instance from a real game, in which Gregory Serper's awareness of this fortress possibility gave him a clear path to a draw against Nakamura in 2004:
[FEN "8/8/8/3k2p1/4p3/2b1P1pP/4KNP1/8 w - - 0 1"]
1.Nxe4! Kxe4 2.Kf1!
Serper heads to the corner with his king, lets the e-pawn go, and will simply play hxg4 if Nakamura ever pushes the g-pawn. It's now clear that we have the essential fortress described above.
Being aware of relatively common fortresses such as this one can be useful in the same way as knowing about standard book draws can be - e.g. bishop plus wrong-color rook pawn is a draw, Philidor's position is a draw in R+P versus R, two knights cannot mate a lone king, etc. At heart, these book draws and fortresses are really all just instances of the same thing: some particular board configuration in which (even a lot of) extra material is useless. (So, knowing that two knights can't mate a king is like knowing about a wonderful fortress you can safely retreat to that doesn't even need any walls.)
In terms of more literal fortresses in which a relatively large amount of material is involved, another point from Valentin's answer is well-taken: widely applicable recipes are few and far between, and "more often than not you have to adapt to the current position." For instance, in Game 9 of the most recent world championship match, Anand as Black succeeded in arranging an appropriate fortress setup against which Gelfand's queen could make no progress against his rook and knight:
[FEN ""]
[Event "Anand-Gelfand World Chess Championship"]
[Site "Moscow RUS"]
[Round "9"]
[Result "1/2-1/2"]
[White "Boris Gelfand"]
[Black "Viswanathan Anand"]
1.d4 Nf6 2.c4 e6 3.Nc3 Bb4 4.e3 O-O 5.Bd3 d5 6.Nf3 c5 7.O-O
dxc4 8.Bxc4 cxd4 9.exd4 b6 10.Bg5 Bb7 11.Qe2 Nbd7 12.Rac1 Rc8
13.Bd3 Bxc3 14.bxc3 Qc7 15.c4 Bxf3 16.Qxf3 Rfe8 17.Rfd1 h6
18.Bh4 Qd6 19.c5 bxc5 20.dxc5 Rxc5 21.Bh7+ Kxh7 22.Rxd6 Rxc1+
23.Rd1 Rec8 24.h3 Ne5 25.Qe2 Ng6 26.Bxf6 gxf6 27.Rxc1 Rxc1+
28.Kh2 Rc7 29.Qb2 Kg7 30.a4 Ne7 31.a5 Nd5 32.a6 Kh7 33.Qd4 f5
34.f4 Rd7 35.Kg3 Kg6 36.Qh8 Nf6 37.Qb8 h5 38.Kh4 Kh6 39.Qb2
Kg6 40.Qc3 Ne4 41.Qc8 Nf6 42.Qb8 Re7 43.g4 hxg4 44.hxg4 fxg4
45.Qe5 Ng8 46.Qg5+ Kh7 47.Qxg4 f6 48.Qg2 Kh8 49.Qe4 Kg7
1/2-1/2
It's not that this was an exact setup Anand already knew about; rather, he was just able to envision this possibility (among others) as the game played out. In his own words:
"My position was worse as he had two bishops for my knights," commented Anand after the game. "My chances hinged on the fortress that I was going to choose. In fact, I had many options and I think I chose the right one."
