circular-economy · materials · manufacturing · family: success's shadow
the engine-block sink is draining
Some 6 million tonnes of wrought aluminium scrap a year is "recycled" by pouring it into engine blocks — and electric vehicles are removing the engine blocks
Problem statement
Aluminium is called "infinitely recyclable," but that is only true within an alloy family. Wrought alloys (car body sheet, extrusions, cans) have narrow compositional limits, and the high silicon content of cast parts "is not admissible" in them; the cast alloys used for engine blocks and gearbox housings are high in silicon, copper and iron (the two main ones, 319 and A380, are designated AlSi6Cu3.5Zn and AlSi9Cu3.5FeZn). Once wrought and cast scrap are shredded together — as end-of-life vehicles are — the mix has traditionally gone "down" into impurity-tolerant cast alloys, because the alloys are hard to separate and because, in the review's words, aluminium melts "can only be refined to a limited extent": the processes that could remove alloying elements and impurity metals "are currently not economically feasible," and impurities such as iron, silicon and copper are "often difficult to remove." Raabe and colleagues estimate that "6.1 Mt of wrought scrap globally are downcycled every year to cast alloys," and that two secondary alloys, 319 and A380, "primarily used in engine blocks and cylinder heads," account for over 80 percent of all die-cast alloy volume and "are currently the most important sinks for scrap." Vehicle electrification removes exactly those parts: "Engine blocks and gearbox housings ... are nowadays made from composition-tolerant recycling alloys, mainly from the aforementioned alloys 319 and A380, for which there is no other relevant use requiring large quantities." The unsolved problem is that the world's aluminium recycling system was quietly balanced on a single sink product that is now disappearing, while the flow of mixed automotive scrap keeps growing.
Why this matters
Remelting scrap uses roughly 5 percent of the energy of primary smelting, so every tonne of scrap that cannot find a home is a tonne of primary aluminium (and, per the review, its "12-16.5 t of GHG per t of metal") that must be made instead. The review reports that mixed automotive scrap supply already exceeds cast-alloy demand, that this "is exacerbated by the future decrease in demand for automotive aluminum castings, forecast due to the growth in electric vehicle sales," and (in an unreferenced sentence following its citation of Hatayama et al. 2012) that "by 2030 there may be an excess supply of > 6 Mt of scrap which is not recyclable due to the prevalence of alloy scrap collected from open-loop recycling. For this scrap the only option at present is dilution by the addition of primary aluminum." Meanwhile, the review says vehicles' wrought-alloy content (sheet, forgings, extrusions) "is strongly increasing" and, citing Zhu et al. 2021, describes a coming wave of automotive body-sheet scrap from high-selling aluminium-sheet vehicles such as the Ford F-150, so the volume of wrought scrap arriving mixed with cast is set to rise just as its sink shrinks. Automakers, secondary smelters and climate policy all depend on this loop continuing to close.
What’s been tried and why it hasn’t worked
The historical answer — downcycle to cast, sweeten with primary — is the thing that is failing. Dilution "is traditionally performed on a limited basis due to the added expense incurred," and the review adds that this "is amplified" by the amount of primary aluminium needed to make scrap from different sources usable in wrought alloys. Sensor-based sorting exists: X-ray transmission separates high-density AlCu/AlZn alloys, and laser-induced breakdown spectroscopy (LIBS) "will discern between a 6010 aluminum alloy and a 6016 alloy," with better results when surfaces are cleaned first. But the authors note that "the scrap sorting technologies used today are well developed, but are reaching their limit due to the large number of alloys and composites on the market," that LIBS spectra "have a certain scatter, as do the sometimes contaminated surfaces of the scrap pieces being probed," and that closely related automotive alloys "if ... not separated ... lose substantial value" — a problem across "the more than 50 million vehicles that reach the ends of their lives every year." Alloy-side fixes have been proposed and tested — Fe/Mn-balanced, Cu-reduced variants of A380, and Krall et al.'s result that alloys with good rollability and with yield strengths and elongations comparable to common automotive sheet alloys can be produced directly from ELV scrap — but the review is blunt about the barrier: such measures "require a willingness of industry and customers to break away from traditional standardization constraints." Rapid-solidification routes such as spray forming can suppress the harmful iron-bearing phases, but the porosity they leave needs further extrusion or forging, which the review says "significantly reduces the attractiveness of the process"; physical removal of iron-rich sludge (sedimentation, centrifugation, filtration) is reported in the literature and, in the review, sized by thermodynamic calculation, and the review says aluminium melts "can only be refined to a limited extent," the alternative refining processes being "currently not economically feasible." In short, the metallurgy of tramp elements is not going to change, sorting is well developed but reaching its limit ("shredder-based recycling practices show natural limits when faced with complex product designs"), and the alloy standards the review calls "traditional standardization constraints" stand in the way of scrap-tolerant alloys.
What would unlock progress
Progress needs a replacement sink or a way to avoid needing one: either (a) new impurity-tolerant "recycling alloys" — for structural castings, EV battery housings, or a Cu-lean, Fe/Mn-balanced 3xx family for non-engine uses — that OEMs and standards bodies actually admit into specifications, or (b) sorting at the shredder good enough that wrought stays wrought (LIBS or hyperspectral classification with machine learning trained on real, dirty, oxidised scrap surfaces rather than clean reference coupons), or (c) dismantling protocols that pull wrought sheet and extrusions off ELVs before shredding, so the cast fraction never contaminates them. The review explicitly points to machine-learning classification of LIBS spectra (Campanella et al.'s neural-network/fuzzy-logic scheme) as the direction where alloy-level sorting has already shown robust recognition. The adjacent precedent is the used-beverage-can loop, where only a few alloys are involved (3004 for the body, 5182 for the lid) and dedicated collection systems reach recycling rates as high as 97 percent in some countries — although the review notes that even the can-to-can cycle "is still not a closed loop." The question is whether a comparable "ELV wrought" stream can be engineered.
Entry points for student teams
A materials/data team could build and benchmark a LIBS-or-hyperspectral alloy classifier on realistically contaminated scrap (oil, paint, oxide, mixed 5xxx/6xxx pieces) and quantify the recognition-rate cliff between clean and dirty surfaces — the gap that decides whether shredder-line sorting is viable. A metallurgy team could take a scrap-tolerant alloy concept from the review (e.g., Cu-reduced A380 with balanced Fe/Mn) and map which non-engine EV components its property window could serve, producing a "sink map" for the post-engine era. A systems team could run a regional material-flow model (one country's ELV fleet, EV share scenarios) to date the year the cast sink closes and size the wrought surplus, then cost dilution versus sorting versus new alloys. Relevant skills: physical metallurgy, spectroscopy, machine learning, material-flow analysis.
Genes
Structural cousins: stuck the same way, in other fields
Sources
Raabe, D., Ponge, D., Uggowitzer, P.J., Roscher, M., Paolantonio, M., Liu, C., Antrekowitsch, H., Kozeschnik, E., Seidmann, D., Gault, B., De Geuser, F., Deschamps, A., Hutchinson, C., Liu, C., Li, Z., Prangnell, P., Robson, J., Shanthraj, P., Vakili, S., Sinclair, C., Bourgeois, L., Pogatscher, S. (2022), "Making sustainable aluminum by recycling scrap: The science of 'dirty' alloys," *Progress in Materials Science* 128, 100947, doi 10.1016/j.pmatsci.2022.100947 — accepted-manuscript PDF (282 pp.) read at accessed 2026-08-18. Supplementary: Krall, P., Weißensteiner, I., Pogatscher, S. (2024), "Recycling aluminum alloys for the automotive industry: Breaking the source-sink paradigm," *Resources, Conservation and Recycling* 202, 107370, doi 10.1016/j.resconrec.2023.107370 (abstract read at the Montanuniversität Leoben research portal record, full text not read), accessed 2026-08-21. go to source 1 ↗ go to source 2 ↗
Working notes
Notes made while writing and checking this brief, shown as written. They use working shorthand.
All quoted language and figures (6.1 Mt/yr wrought-to-cast downcycling from 2012 data; alloys 319/A380 >80 percent of die-cast volume; the ">6 Mt by 2030" surplus estimate, which the review states without a reference of its own immediately after citing Hatayama et al. 2012 [ref. 168] for the EV-driven fall in casting demand — the attribution to Hatayama is therefore inferred from placement, not explicit; the ~5 percent remelt-energy figure; ">50 million vehicles" per year; the LIBS 6010/6016 statement) are from the Raabe et al. accepted manuscript read in full-text PDF on 2026-08-18. The Krall et al. 2024 abstract (SSRN/ScienceDirect were HTTP 403; abstract text taken from search-engine and OUCI index summaries) supports the claim that direct ELV-scrap-to-wrought-sheet is feasible at lab scale — should be confirmed against the paper. The 12–16.5 t GHG/t primary figure and the ~5 percent remelt-energy figure are both from the review's introduction. The claim that aluminium-intensive vehicles are 'beginning to reach end of life' is the author's inference from Zhu et al. 2021 (cited by the review as ref. 36, 'The coming wave of aluminum sheet scrap'), whose full text was not read. `failure:success-caused` was applied after running the taxonomy test: the intervention (drivetrain electrification) achieved its stated objective; the harm (loss of the only large impurity-tolerant sink for mixed scrap) follows through a specific, named mechanism (no engine blocks/gearbox housings = no demand for 319/A380); and the harm is structurally coupled to the success (capability-displacement / infrastructure-lag sub-type). `failure:wrong-problem` was ruled out (downcycling was a right-enough objective while the sink existed). `temporal:worsening` passes all three requirements: mechanism (growing EV share removes the sink while ELV aluminium content rises), trajectory (supply already exceeds cast demand; >6 Mt surplus projected by 2030), and barrier — not merely salience — worsening. `constraint:installed-base` covers the alloy-standard lock-in the review names ("traditional standardization constraints"). `stakeholders:multi-institution` passes: shredders/scrap dealers own sorting, remelters own alloy production, OEMs and standards bodies own the specifications a new recycling alloy must enter. `constraint:coordination` was considered and rejected on filter (2): the binding constraints are metallurgical (tramp elements) and sorting performance, not unwillingness to cooperate. Related collection briefs: `energy-lfp-battery-recycling-economics` (a different EV-driven end-of-life problem), `manufacturing-reuse-quality-standardization` (grading reused parts, not alloy sorting); no existing brief covers aluminium alloy sorting or the cast-alloy sink.
Source type: Self-articulated (materials-science community review naming the sink problem)
Verified at intake 2026-08-18: gate (net) + adversarial source check + contested-tag second coding.
Reconciliation 2026-08-21: Queued by the thin-source lint (single-URL Source line with ≥8 unit-bearing figures), not by a content flag. The Raabe et al. accepted-manuscript PDF was re-downloaded from the cited dierk-raabe.com URL and read in full text (282 pp., pdftotext) on 2026-08-21; every unit-bearing figure and quoted phrase in the brief is present verbatim in it: "6.1 Mt of wrought scrap globally are downcycled every year to cast alloys" (from 2012 data, ref. 8); 319/A380 "usage volume equals over 80% of all die-cast alloys," "primarily used in engine blocks and cylinder heads," "currently the most important sinks for scrap"; the engine-block/gearbox-housing sentence ending "no other relevant use requiring large quantities"; the "> 6 Mt" by-2030 surplus sentence (confirmed unreferenced and immediately following the ref.-168 EV-forecast sentence, exactly as the intake note states); "12-16.5 t of GHG per t of metal produced"; the ~5 percent remelt-energy figure; "more than 50 million vehicles that reach the ends of their lives every year"; the LIBS 6010/6016 sentence; the spectra-scatter, sorting-technologies-"reaching their limit," "lose substantial value," dilution-"added expense incurred," and "traditional standardization constraints" quotes; XRT currently used to separate AlCu and AlZn alloys; and the Campanella et al. neural-network/fuzzy-logic classification scheme (ref. 770). The one intake caveat — Krall et al. 2024 known only from search-engine/index summaries — is now closed: the abstract was read on the Montanuniversität Leoben research portal record (https://pure.unileoben.ac.at/en/publications/recycling-aluminum-alloys-for-the-automotive-industry-breaking-th), which confirms the brief's claim ("alloys with promising properties such as good rollability, yield strengths and elongations comparable to common automotive sheet alloys can be produced directly from ELV scrap") and supplies the volume number (202), now added to the Source line; Crossref metadata for doi 10.1016/j.resconrec.2023.107370 matches (Krall, Weißensteiner, Pogatscher; RCR 202, 107370, March 2024). No factual changes: the brief legitimately rests on the single tier-1 review for its figures, and every one was verified in-source 2026-08-21.
Reconciliation 2026-10-01: Fact repair after reading the whole write-up against the sources it names; a second reader went over the repairs on the same day, re-fetched the same sources and reworded two of them (items 2 and 7 below describe the final wording). The Raabe et al. accepted manuscript was fetched on 2026-10-01 from the dierk-raabe.com address on the Source line (PDF, 282 pages) and its full text was searched for every figure, name and quotation. For Krall et al. 2024 only the abstract could be read, on the Montanuniversität Leoben portal record on the Source line (fetched 2026-10-01); the publisher page is closed to scripts and no open full-text copy was found. Both citations were compared with their Crossref records (api.crossref.org, by DOI) and match. Zhu et al. 2021, named in the notes above, could not be read (closed access, no abstract in the open record); only what the review says about it was used. (1) The text put the words "primarily in engine blocks and cylinder heads" in quotation marks. The review's words are "primarily used in engine blocks and cylinder heads"; the quotation now follows the review. (2) The text said wrought alloys "tolerate almost no silicon," that cast alloys "contain 6–10 percent silicon," and that iron and silicon "cannot practically be refined back out." The review does not give those figures or that absolute statement. It says wrought alloys have narrow compositional limits, that the high silicon content from cast parts "is not admissible in wrought alloys," that the engine-block alloys 319 and A380 are AlSi6Cu3.5Zn and AlSi9Cu3.5FeZn, that automotive scrap has been downcycled because the alloys are difficult to separate, that aluminium melts "can only be refined to a limited extent" and the alternative refining processes that could remove alloying elements and impurity metals "are currently not economically feasible," and that iron, silicon, copper and tin are "often difficult to remove." The sentences now say that. (3) The text said the first generation of aluminium-intensive vehicles, including "EV battery enclosures and structural castings," is beginning to reach end of life. The review does not say this. It says vehicles' wrought-alloy content "is strongly increasing" and shows, from Zhu et al., a large coming growth in body-sheet scrap from the Ford F-150 and three other high-selling models; the sentence now says that. The older note above that calls this claim an inference from the title of Zhu et al. refers to the wording that has now been replaced. (4) The text said the economics of dilution "get worse as the fraction of mixed scrap rises." The review says the expense "is amplified" by the amount of primary aluminium needed to make scrap from different sources usable in wrought alloys; the sentence now says that. (5) The text said LIBS tells 6010 from 6016 "on clean surfaces." The review says the technique does this and that results improve when surfaces are cleaned first; the sentence now says that. (6) The text said alloy fixes were "demonstrated at lab scale" and described the Krall et al. result as "cold-rolled sheets made directly from simulated ELV scrap." Neither phrase is in the review or in the Krall abstract. The abstract says alloys with good rollability, and yield strengths and elongations comparable to common automotive sheet alloys, can be produced directly from ELV scrap; the sentence now says that. The clause that such alloys "must still be qualified into products" was not in either source and was removed. (7) The text said physical impurity removal "(spray forming, rapid solidification, filtration) works on iron intermetallics only at high cost." The review treats spray forming as a rapid-solidification manufacturing route, not as impurity removal, and gives its drawback as porosity that needs later extrusion or forging, not cost; it reports sludge removal by sedimentation, centrifugation or filtration from the literature and sizes it by thermodynamic calculation; and it says that aluminium melts "can only be refined to a limited extent" and that the alternative refining processes are "currently not economically feasible." The sentence now follows the review. The closing summary of that section was trimmed to match ("sorting is well developed but reaching its limit," with the review's own words that "shredder-based recycling practices show natural limits when faced with complex product designs"; the statement that alloy standards "were written for primary metal" and the phrase "shredder scale and speed" are not in the sources and were replaced by the review's "traditional standardization constraints"). (8) The text said the beverage-can loop is "collected separately and closed at ~can-to-can quality." The review says few alloys are involved (3004 for the body, 5182 for the lid), that countries with dedicated collection reach recycling rates as high as 97 percent, and that the can-to-can cycle "is still not a closed loop"; the sentence now says that. Confirmed in the review and left as written: 6.1 Mt of wrought scrap downcycled to cast alloys each year (from 2012 data); 319 and A380 at over 80 percent of die-cast alloys and "currently the most important sinks for scrap"; the engine-block and gearbox-housing sentence; "12-16.5 t of GHG per t of metal"; remelting at 5 percent of the energy of primary production; supply of mixed automotive scrap exceeding demand; the "> 6 Mt" by 2030 sentence and its placement; the dilution, sorting-limit, spectra-scatter, "lose substantial value," 50 million vehicles and "standardization constraints" quotations; X-ray transmission for AlCu and AlZn alloys; the Campanella et al. scheme. Not checked: anything in Krall et al. beyond its abstract, and Zhu et al. itself. The long title's "pouring it into engine blocks" is shorthand: the review says the 6.1 Mt goes to cast alloys, of which the engine-block alloys are the largest part; the title was left as it is. No title changed. Genome tags untouched.