DGF15/25B Cryogenic Strainer

| Application Overview

The DGF15/25B is the second branch of the DGF product line — its name alone reveals its positioning: the suffix letter B stands for Basket, not the dual-drum structure of the preceding DGF15/25. This is a fundamentally different design philosophy: the DGF is inline — medium passes through the dual-drum filter cartridge along the axis, with the outlet on the same line; the DGF-B is Y-type — medium enters the body and encounters an approximately 45° angled branch chamber, the filter basket sits submerged in this angled cavity, and medium passes through the basket wall before redirecting back toward the main flow path and exiting.

This Y-type layout brings three direct consequences. First, height is dramatically compressed. DGF-15's H=53 mm; DGF-15B's H=39 mm — 14 mm shorter (about 26%). In the tight installation space of a tank-bottom valve cluster, this height difference is the divide between "fits" and "doesn't fit." Second, basket maintenance requires no pipe disconnection. The DGF is an inline design — cleaning the filter requires removing the entire strainer from the pipeline; butt-weld ends make this nearly non-serviceable. The DGF-B only needs the two M bolts on top loosened, the flange cover and basket lifted out, cleaned, and reinstalled — the piping stays untouched. Third, the Y-type chamber brings its own settling effect. After medium enters the angled chamber, velocity drops due to the sudden cross-sectional expansion; particles denser than LN2/LO2 (metal chips, weld spatter) plunge toward the basket bottom under inertia — the basket wall intercepts fine particles, gravity settling captures coarse particles, with both mechanisms operating in parallel.

Another major change in DGF-B is the connection: from DGF's butt weld / socket weld to NPT tapered pipe thread — DGF-15B uses NPT 1/2, DGF-25B uses NPT 1. This is not a downgrade but an orientation toward a completely different installation scenario: the tube system of tank. Tanks leave the factory with abundant NPT female-threaded ports — safety valve ports, level gauge ports, pressure gauge ports, fill/discharge ports. DGF-B can screw directly into these reserved ports — no welding, no flange adapters, no extra pipe fitting.

But NPT threads also bring a pressure downgrade: DGF-B's rated pressure is 40 bar / 580 PSI, lower than DGF's 50 bar / 725 PSI. This is not a limitation of the body itself — the body remains the same grade of CF8 stainless steel casting — but the sealing capacity ceiling of the NPT tapered pipe thread joint. NPT's sealing principle is thread-flank interference fit (1:16 taper), with metal-to-metal contact between thread crests and roots forming the seal. This sealing mechanism is reliable to the 600 PSI range at room temperature, but at −196°C, the male and female threads contract at different rates and different speeds — the male thread (external) contracts faster and earlier than the female thread (internal) — thread interference may recede below the sealing threshold at cryogenic temperature, causing micro-leakage. The 40 bar rating has already built in ample safety margin for this cryogenic thread-sealing uncertainty.

The parts list also reveals a key difference in the B-type: one additional part — a PTFE gasket (item 3), sandwiched between the flange (item 4) and the body (item 1). The DGF's all-stainless-contact faces rely on machining precision for metal-to-metal sealing; after each disassembly, the sealing faces need flatness inspection. The DGF-B uses a PTFE gasket as a one-shot sealing solution — after opening to clean the basket, if the gasket shows no permanent deformation, it can be reused; if a permanent impression has been set, a fresh gasket restores sealing in seconds, with no re-machining of the flange faces required. PTFE hardens at −196°C, but as long as the gasket groove design provides sufficient compression allowance (PTFE's thermal contraction at low temperature far exceeds that of stainless steel), sealing remains reliable.

| Working Principle

Fluid dynamics of the Y-type chamber: not "a bend," but a combined diffusion-settling-filtration cavity

The DGF-B's core design is a "Y-type" body — externally, the inlet and outlet lie on the same straight line, but a roughly 45° branch chamber opens in the middle, and the filter basket sits inside this branch chamber. This geometry creates a three-stage flow path:

Stage one — inlet acceleration zone. Medium enters from the left NPT port, with the passage cross-section equal to the pipe bore. Velocity is at pipe-flow level (typical LNG offloading line velocity is 3~5 m/s).

Stage two — diffusion-settling zone. When medium reaches the Y-chamber entrance, the flow cross-section suddenly expands to the branch chamber's area — this area is far larger than the pipe cross-section. By the continuity equation (A₁v₁ = A₂v₂), velocity drops to a fraction of its original value in this step. With velocity reduced, solid particles denser than the medium (SS debris density ~7.8 g/cm³, LNG density ~0.45 g/cm³, density ratio ~17:1) lose the fluid's carrying capacity — they can no longer be "carried along" by the flowing LNG, but plunge toward the basket bottom under gravity and inertia.

But there is a temperature factor working in favor here: LNG's viscosity is extremely low (~0.12 mPa·s, 1/8 of room-temperature water). Low viscosity means particles settle faster in the fluid — in Stokes' settling equation v = (2/9) × (ρₚ−ρ_f) × g × r² / μ, the denominator μ being tiny means settling velocity v is enormous. The same 100 μm SS particle settles approximately 8× faster in LNG than in room-temperature water. This is why the Y-chamber achieves efficient filtration in cryogenic service despite having no spring compensation or dual drums: the medium's inherent low-viscosity properties make gravity settling the first effective filtration barrier.

Stage three — basket-penetration confluence zone. After medium penetrates the basket wall and enters the basket's central cavity, it reconverges into the outlet passage. The outlet cross-section matches the inlet, velocity recovers to pipe-flow level, and filtration is complete.

The cryogenic sealing game of NPT threads: the taper is both advantage and risk

NPT thread sealing does not rely on gaskets, adhesives, or PTFE tape — it relies on the thread itself: the thread taper is 1:16 (diameter changes by one unit for every 16 units of axial advance). When the external thread screws into the internal thread, an interference fit develops between thread crests and roots, forming a continuous metal-to-metal contact band distributed along the thread helix. Theoretically, as long as interference is sufficient, this contact band is sealed.

But at cryogenic temperature, this "theory" is shaken by two physical facts:

First, male and female threads contract with a time lag. When −196°C medium flows through the strainer, the innermost NPT male thread (the stub connected to the body) contacts the cryogenic medium first — its temperature drops to near −196°C within seconds, and it contracts. But the mating NPT female thread on the tank port is on the far side of the tank nozzle flange — the tank nozzle is a large-mass heat sink, and the cold takes conduction time to travel from the inner pipe wall to the outer wall. During this time-lag window, the male thread has already contracted by about 0.2%~0.3% (SS linear contraction over a 196°C temperature swing), while the female thread has barely contracted — thread interference temporarily diminishes or vanishes. If the system is already pressurized at this point (e.g., during pre-cooling), this time-lag gap becomes a leakage channel.

Second, NPT thread interference diminishes with repeated make-break cycles. Each screw-in/screw-out cycle causes minute plastic deformation of the thread crests — crest metal is compressed and flattened. After three make-break cycles, interference may drop from the factory 50100 μm to below 2030 μm — this margin still seals at room temperature, but at cryogenic temperature, the male-female contraction differential (0.020.05 mm for NPT 1/2 grade) could consume the entire remaining interference.

This is why the DGF-B's pressure rating is set at 40 bar rather than 50 bar (the DGF's weld-end rating): not because the body is weaker, but because the NPT joint's sealing uncertainty under the combined cryogenic + make-break-history scenario requires 10 bar of derating margin. If installed and never removed, the DGF-B's NPT joint could theoretically handle near DGF's 50 bar — but the strainer is designed for periodic disassembly to clean the basket; "never removed" is unrealistic.

The PTFE gasket: an elegant solution to the "re-seal after disassembly" problem

The DGF-B's flange-to-body sealing face uses a PTFE gasket (item 3). PTFE over spiral-wound metal or PEEK gaskets — three reasons:

Replaceability. The basket needs periodic removal for cleaning; each disassembly requires re-tightening the flange. Metal-to-metal sealing degrades after multiple disassemblies due to particle embedding or scratches — the sealing faces need re-lapping to restore. The PTFE gasket is designed as a consumable: if the gasket shows permanent deformation upon disassembly, insert a fresh one, and sealing is restored as-new.

Cryogenic conformability. PTFE hardens at −196°C but retains faint plasticity — its yield strength at low temperature is approximately 3040 MPa, less than 1/10 of stainless steel. When flange bolts are tightened, PTFE undergoes controlled plastic flow, filling microscopic unevenness on the flange face (machining marks at Ra 1.63.2 μm level) and forming conformal sealing. A metal gasket cannot do this — micron-scale surface unevenness becomes a leakage path.

Contraction compensation. PTFE's thermal contraction from 20°C to −196°C is approximately 1.5%~2% (far greater than SS at 0.3%). If the gasket groove is designed deep enough (e.g., 2 mm deep, gasket 2.2 mm thick, 0.2 mm compression), the PTFE thickness reduction after cold-shrink (2 mm × 1.5% = 0.03 mm) is far less than the initial compression (0.2 mm), and bolt preload still maintains sufficient gasket contact stress.*

|

 Factory Shoot

|

 Application Video

|

 Product parameters

| Data Sheet

|

Comment & FAQ

1: What does the B suffix mean on DGF15/25B? How to choose between DGF and DGF-B?

A: B stands for Basket, denoting the Y-type single-basket design. The selection decision boils down to two questions: DGF15/25 for line engineers — permanent weld, infrequent removal, pressure traded for sealing integrity. DGF15/25B for maintenance engineers — screw-in, five-minute basket clean-out, pressure traded for convenience.

2: Are NPT threads reliable at cryogenic temperatures? Any leak reports from users?

A: NPT thread reliability at cryogenic temperature depends on three variables — tightening torque, make-break count, and PTFE tape use — any one out of control leads to leakage. Most reliable approach (recommended): on initial installation, wrap 23 layers of PTFE tape on the NPT threads (choose oxygen-compatible tape per MIL-T-27730A; do not use ordinary plumbing tape — it contains oil, hazardous in LO2), tighten with a torque wrench to ASME B1.20.1 standard torque (NPT 1/2 ≈ 4055 N·m, NPT 1 ≈ 70~90 N·m). The tape's role is not primary sealing — primary sealing remains the thread's metal-to-metal interference — the tape is "backup sealing," filling microscopic scratches and thread-crest micro-notches that may form during make-up. If metal interference temporarily recedes at cryogenic temperature, the tape still fills the thread gap. Least reliable approach: dry make-up (no tape), torque by feel, reuse old tape after multiple disassemblies. After three make-break cycles, the tape is shredded and thread interference is degraded; leakage is only a matter of time. Real-world data: in LNG fueling station tank NPT joint inspections, micro-leaks were almost entirely concentrated on joints disassembled 3+ times — while joints installed once and never touched had near-zero leak reports. The rule is simple: if you know this strainer will be frequently opened for cleaning, replace the PTFE gasket and PTFE tape with fresh ones every disassembly. Do not cut costs on these two consumables — their unit price is less than 1% of the strainer's purchase cost, but their failure renders the entire strainer useless.

3 :  Will the PTFE gasket crack at cryogenic temperature? Is it more reliable than metal gaskets?

A: PTFE will not crack — its embrittlement temperature is −268°C (near absolute zero), far below the DGF-B's −196°C operating floor. However, PTFE loses its room-temperature softness at low temperature — its elastic modulus rises from 0.5 GPa at room temperature to 23 GPa at −196°C (hardening by about 46×), making it no longer a "soft" gasket but closer to a "hard" engineering plastic. This brings a design pitfall: if the gasket groove depth is designed for "PTFE compression at room temperature" without accounting for cryogenic hardening, then after cooldown the hardened gasket can no longer continue compensating for flange-face stress relaxation — once bolt preload is "locked in" by the hardened gasket, if the flange faces undergo micro-deformation during temperature cycling, the gasket cannot plastically flow and fill as it does at room temperature, and leakage opens. Correct design: the gasket groove depth should cause the PTFE gasket to undergo 15%~25% thickness compression at room-temperature tightening (not 5%~10% light compression); this compression reserve maintains sufficient contact stress after cryogenic hardening. Compared to spiral-wound gaskets: spiral-wound (e.g., 304+graphite) has higher cryogenic sealing reliability than pure PTFE, but spiral-wound gaskets are not reusable — must be replaced every disassembly — and require higher bolt preload (>50 MPa gasket stress typically). PTFE gaskets can be reused within 3 disassemblies (if no permanent deformation) with lower bolt torque requirements — for a strainer that needs frequent opening for basket cleaning, PTFE has lower total cost of ownership.

4:  How often to clean the basket? How to tell it's clogged? Same method as DGF?

A: The judgment method differs — DGF-B cannot use the DGF's "differential pressure method" as the primary means, because the DGF-B is mounted on tank ports, and most installations lack upstream/downstream pressure taps. Alternative judgment methods (in priority order): 1. Flow observation (most direct): during normal fill/discharge, if the pump discharge pressure is unchanged but the flowmeter reading drops by 15%+ — the basket is likely clogged. Principle: basket clogs → local flow resistance increases → the pump's operating point shifts up its characteristic curve → flow drops at the same discharge pressure. 2. Temperature anomaly (auxiliary): if the basket is severely clogged, the strainer body temperature will be higher than normal — because the Joule-Thomson effect creates localized pressure-drop/temperature-rise at the clog. Hold the back of your hand near the strainer body (do not touch directly — cryogenic burn risk); if the DGF-B's frost coverage extends beyond normal operation range (particularly around the Y-chamber where the basket sits), it indicates abnormal internal throttling. 3. Time-baseline method (preventive): don't wait for clogging. Record the intervals of the first three cleanings, take the average as a fixed cleaning cycle. In typical LNG fill/discharge lines, the DGF-B basket needs its first inspection after 3~6 months of continuous operation — the cleaning interval depends on line cleanliness. Practical tip: after the first installation, open and inspect the basket once within one month — this reveals the line's "contamination rate." If after one month the basket has only minor debris (<10% area covered), extend the next inspection to three months. If after one month the basket is heavily covered, it indicates abundant welding residue — set the cleaning cycle at one month until three consecutive cleanings show debris quantity stabilizing, then extend.

5:  Does a Y-type strainer have higher pressure drop than an inline strainer? Does it affect pump suction performance? 

A: Quite the opposite — a Y-type strainer's pressure drop is typically lower than an inline strainer of the same bore. The reason is the Y-chamber's "diffusion-basket-penetration" flow path: medium entering the large-cross-section Y-chamber drops in velocity, and the local velocity through the basket wall is far lower than in an inline type (where medium impacts the screen surface directly at full pipe velocity). Lower basket-penetration velocity = lower screen local resistance loss. Comparative estimate for DGF-15B vs DGF-15: clean basket (100 mesh), LNG medium, rated flow — DGF-15B pressure drop 0.30.8 PSI, DGF-15 pressure drop 0.51.5 PSI. DGF-B's pressure-drop advantage comes from the improved basket-surface-area-to-pipe-cross-section ratio — the Y-type basket's external surface area far exceeds the inline type's flat-screen/drum area. Impact on pump suction: NPSH (Net Positive Suction Head) is extremely sensitive to strainer pressure drop — every 1 PSI of strainer pressure drop equals approximately 1.6 meters of LNG NPSH loss. If the pump's NPSH margin is only 2~3 meters to begin with, the strainer's 0.5-meter added pressure drop could be the dividing line between "no cavitation" and "cavitation." Therefore, for pump suction lines, prefer Y-type (low pressure drop) over inline (high pressure drop) — if the DGF-B's bore and connection type happen to match the pump suction line.